An E>L.Amination of the Concept of Arousal Within the Context of The

{

An E>l.amination of the concept of Arousal within the Context of the Sexual Behaviour of the snail, Helix aspersa

Shelley A. Adamo

Department of Biology McGill university, Montreal

April, 1990

A thesis submitted te the Faculty of Graduate Studies and Research of McGill University in partial fulfillment of the

requirements for the degree of Ph.D.

@shelley Adamo, 1.990 .------._--_.

2 (

Ta my parents, with love and gratitude for aIl their

support and encouragement 3 Abstract

Sexual 'arousal' in Helix aspersa can be divided into 2 components, sexual proclivity (the tendency of a snail to respond to conspecific contact with courtship) and sexual arousal (the intensity with which the snail courts) . Sexual proclivity and sexual arousal have different effects on feeding and locomotion and are differentially affected by sexual isolation, daily conspecific contact, and by a courtship pheromone found in the digitiform gland mucus. Therefore sexual arousal and sexual proclivity ara probably mediated by 2 separate ph}rsiological mechanisms. Behavioural state, or the animal' s general level of activity, correlates positively with mating behaviour. However, al though a central system controlling behavioural state probably exists, it has no direct effect on either sexual proclivity or sexual arousal. Confusion over the term 'arousal', which impedes neuroethological research in this area, would be decreased by the adoption of the terms used in this thesis. 4 Résumé "L'éveil" sexuel chez Hf'lix aspersa peut être divisé en deux composantes: la disposition sexuelle (soit la

tendance d'un escargot à faire la cour lors d'un contact avec un congénère), et la stimulation ou l'éveil seyuel

proprement dit (c'est-à-dire l'intensité avec laquelle l'escargot fait la cour). Non seulement la disposition et l'éveil sexuels ont-ils des effets distincts sur les comportements d'alimentation et de locomotion, mais ils sont affectés de manière différente par l'isolation sexuelle, le contact quotidien avec les congénères 1 et par une phéromone rel iée au comportement de cour que l'on trouve dans le mucus des glandes multifides. Il est donc fort probable que la disposition sexuelle et l'éveil sexuel soient servis par deux mécanismes physiologiquement distincts. L' étât comportementa l (l e ni veau général d' activi té de l'animal), est en corrélation positive avec le comportement d' appariement. Cependant, bien qu'il soit possib.le qu'un système central contrôlant l' étât comportemental existe, celui-ci s'avère sans effet, autant sur la disposition sexuelle que sur l'éveil sexuel. La confusion règnante en ce qui concerne le terme "éveil", et qui de fait freine la recherche en neuroéthologie dans ce domaine, se trouverait diminuée si l'on adoptait les termes utilisés dans cette thèse. 5 Preface

The research presented in this thesis, and the preparation of the thesis i tself, are solely the work

and responsibil i ty of the candidate. However, in the

experiments requiring blind contraIs, Barbara Tolloczko,

Caroline Tidd and Shelley LaBerge assisted with the data

collection.

The thesis chapters which have been publ ished, or

have been submitted for pUblication, are listed below:

Chapter 2. Adamo, S.A. and Chase, R. (1988), Canadian Journal of Zoology, 66: 1446-1453.

Chapter 3. Adamo, S.A. and Chase, R. (in press),

Journal of Experimental Zoology.

Chapter 4. Adamo, S.A. and Chase, R. (in press),

Behavioral and Neural Biology.

Chapter 5. Adamo, S.A. and Chase, R. (in press),

Behavioural and Neural Biology.

Chapter 6. Adamo, S.A. and Chase, R. (submitted),

Behavioural and Neural Biology.

Chapters 2, 3, 4, 5, and 6 are in the formats

required of them by their respective journals.

Chapter 1 gi ves a general introduction to the topic

and outlines the scientific problem addressed in this

thesis. The discussion (Chapter 7) surnmarizes the

resu1 ts of the thesis in the context of the scient i fic ," question posed in the introduction. 6 ( Contributions to Original Knowl.eoge

Below is a listing of the main contributions to original knowledge contained in this thesis: 1) The thesÜ; gives a clear description of Helix aspersa sexual behaviour and copulation (Chapter 2) which previously had been poorly characterized.

:2) It presents the f irst known function for dart shooting behaviour (Chapter 3) . It demonstrates that the digitiform gland mucus, which coats the dart, contains a courtship pheromone.

3) It establishes that the phenomenon of 'arousal' can

be divided into 2 distinct components, arousal and proclivity (Chapter 4, Chapter 5) and that this division can be usefully applied to other molluscs (Chapter 7).

4) It demonstrates that the relationship between sexual

behaviour and feeding behaviour in Helix can best be described as a time-sharing arrangement (Chapter 5). This is the first time that this complex behavioural interaction has been shown in a mollusc.

5) It clarifies the meaning of the term central

1 arousal' (Chapter 6). Chapter 7 shows that much of the controversy in the literature dealing with central

1 arousal' is due to a confusion in tenninology.

clarify ing the term 1 arousal' will aid neuroethologists in their study of motivated behaviours. 7

The following statement is added in accordance with the regulations of the Facul ty of Graduate Studies and Research of McGill University: The candidate has the option, subject to the approval of the department, of including as part of the thesis the text, or duplicated published text (see below), of an original paper, or papers. In this cas~ the thesis must still conform ta all other requirements explained in Guidelines Concerning Thesis preparation. Additional material (procedural and design data as well as descriptions of equipment) must be provided in sufficient detail (e.g. in appendices) to allow a clear and precise j udgemen t to be made 0 f the importance and originality of the research reported. The thesis should be more than a mere collection of manuscripts published or to be published. It must include a general abstract. a full introduction. and literature review and a final overall conclusion. Connecting texts which provide logical bridges between different manuscripts are usually desirable in the interests of cohesion. It is acceptable for thesis to include as chapters authentic copies of papers already published, provided these are duplicated clearly on regulation thesis stationery and bound as an integral part of the thesis. Photographs or other materials which do not duplicate well must be included in their original forme In such instances. connecting texts are mandatory and supplementary explanatory material is almost always necessary.

The inclusion of manuscripts co-authored by the candidate and ethers is acceptable but the candidate is required te rnake an explici t statement on who contributed to su ch work and to what extent, and supervisors must attest to the accuracy of the claims, e.g. before the Oral Committee. Since the task of the Examiners is made more difficult in these cases, i t is in the candidate' s interest to make the responsibilities of authors perfectly clear. Candidates fellowinq this option must inform the department before it submits the thesis for review. { 8 Acknowledgements Although it is established etiq.lette to thank one's supervisor for his time and expertise, my thanks to Dr. Ronald Chase are sincerely warm and heart-fel t. He has

always been support ive of me and my research. His keen ability to analyze, and his own integrity and imagination as a scientist, has provided me with an invaluable learninq experience. I would also like to thank the members of my supervi::;ory committee, Dr. Valerie Pasztor, Dr. Gerry Pollack, and Or. Henry Reiswiq for their encouragement and constructive criticisms. l would like to thank Robert Lamarche and Guy L' Heureux for their advice and expert technical assistance with the photographie materials and with the laser printer. l would also like to thank Matthias Schmidt for his clear and literate translations of Meisenbeimer (1912) and Syzmanski (1913).

l thank the Natural Sciences and Engineering Research Council of Canada for their financial support. l would also like to thank my colleagues in the Chase

Lab., especially Holly Bradshaw and Barbara Tolloczko, for their constant encouragement and for their qifts of both time and expertise. Finally l would like to thank my relatives and friends, especially my parents and Jim Belanqer, who { were always willing to listen. - 9 Table of Contents

Page Abstract ...... 3

Résumé •..••••.•••...•••..•••••..••••..••••...••. 4

Preface ...... 5

Acknowledqements •...•••...•••..•.•••...••.....•• 8

Chapter 1. Thesis Introduction ..••....•.....•.. 10

Chapt.er 2. Courtship and Copulation in the Terrestrial snail, Helix aspersa 31

Chapter 3. The' Love Dart 1 of the Snail, Helix aspersa Inj ects a Pheromone that Decreases Courtship Duration •....•.. 52

Chapter 4. Dissociation of Sexual Arousal and Sexual Proclivity in the Garden Snail, Hel ix aspersa ...... 75 Chapter 5. The Interactions of Sexual Behaviour, Feeding Behaviour and Locomotion in the Behaviourt.l Hierarchy of the

Snail l Hel ix aspersa .••....••...... 98

Chapter 6. 1 Central Arousal 1 and Sexual Responsiveness in the snail, Helix

jJlpersa ..•....••.•...••...•••.....•. 122

Chapter 7. Thesis Discussion •..••••..••••...••. 155

Bibl iography ...... 181

, ,~ { 10

Chapter 1. Thesis Introduction

{ 11

Introductory remarks on the snail« Helix aspersa

Helix aspersa is a terrestrial snail found in many temperate parts of the world (Herzberg and Herzberg, 1962) . Its distribution is restricted by its need for high humidity and moderate temperatures. Possibly due to its sensitivity to dry air, H. aspersa tends to be most active at night. Snails are capable of aestiva":ing when conditions become adverse. A snail will withdraw into its shell and seal the opening with a membrane to retard moisture loss.

By aestivating when conditions are poor, snails can survive for several years, reproducing annually (Potts,

1975; Bailey, 1981). !L....-.aspersa is a general herbivore, grazing on a wide val: iety of plants (Ingram, 1946). In California its successful foraqing abilities have led it: to become a major agricultural pest (Ingram, 1946). Under laboratory conditions, snails will aiso feed on dead conspecifics (personal observations), and snails probably feed on dead animal tissue in the field as weIl. Like aIl molluscs, Helix relies on a hydrostatic skeleton for the support of its body l'arts against gravity (Dale, 1973). The snail has a semi-open circulatory system with arteries leading blood away from 12 the heart. Blood returns to the heart via hemocoels, which are large, open, blood sinuses (Dale, 1973). During emergence from the shell, increases in blood pressure are needed t.o 1 inflate 1 the body cavity before any active moveme!1t can occur. Blood pressure can be increased by constricting arteries to shunt blood flow, and, as is do ne during dart shooting, by partially contracting the body wall (Dale, 1973; chung, 1986). Increases in heart rate, as well as in the amplitude of its contraction, may be invol ved as weIl (Lehman and

Greenberg, 1987).

Coordinating the activity of the circulatory system with the rest of the snail, are the central and peripheral nervous systems. Molluscs have extensive peripheral nerve nets which have their own cell somata (Kandel, 1979). General coordination, however, is thought to be executed by the central nervous system.

The central nervous system consists of a cerebral ganglion, which sits dorsal to the crop, as weIl as a series of 7 fused ganglia which are located immediately ventral to the crop. As in other mOlluscs, the central nervous svstem is amenable to neurophysiological analysis because it has relatively large cells, many of which are uniquely identifiable and have already heen c'laractérized

(Pentreath et. al., 1982). This has made Helix, and other mOlluscs, notably Aplysia, popular as model systems 13 for the exploration of the biological basis of complex behaviours (Chase, 1986; Kupferrnann, 1974).

Sorne of the most complicated behaviours performed by the snail are those involved in reproduction.

H. aspersa reproductive biology

H. aspersa is a simultaneously reciprocal

hermaphrodi te. Each snail has both male and female

geni tal organs and an ovotestis which makes both sperm

and eggs. It cannot self-fertilizei it must have access

to foreign sperm for successful reproduction (Tompa,

1984) .

The following paragraphs give a brief summary of Helix

aspersa reproductive physiology. Fig. 1 in Chapter 2 gives a schematic outline of its reproductive organs.

Snails package their sperm in spermatophores prior to

copulation. During copulation, each snail everts i ts

penis and pushes it into its partner's genital atrium and bursa tract. The penis has a bulbous hook at one end to maintain the connection during spermatophore transfer.

The spermatophore enters the bursa tract diverticulum and slowly dissolves, releasing and activating the sperm

(personal observations; Lind, 1973).

The sperm swim down the tail of the spermatophore, back into the genital atrium, and then up the 14 spermoviduct. Sperm are stored in a fertilization pou ch at the end of the spermoviduct (Lind, 1973). Ovulation does not necessarily follow copulation and the 2 activities appear to be independently controlled. During ovulation, ova mature in the ovotestis and then pass through the hermaphroditic duct which conducts them to the fertilization pouch. Eggs are fertilized and then coated with a nutrient material made by the albumen gland. The eggs pass through the spermov iduct and are extruded out the genital atrium (Tompa, 1984). snails require an average of 9 hours to complete egglay ing (Herzberg and Herzberg, 1962). Snails tend to be slow at completing aIl reproductive behaviours; copulation requires about 7 hours and courtship usually requires more than 30 min (Chapter 2). Because sexual behaviour requires a substantial time investment, the behaviour is likely to be ecologically costly (Peake, 1978). Nonetheless, sexual behaviour has evolved yet another feature, dart shooting, which adds an additional cost to reproduction. In one of the most violent precopulatory behaviours known in the animal world, snails impale their partners wi th a 1 cm calcareous dart. The dart is propelled by the eversion of the muscular dart sac (Meisenheimer, 1912), and is coated by a mucus secreted by the digitiform glands (Chung, 1986). 15 Malacologists have speculated on the purpose of dart shooting for over a century (Tryon, 1882), ,but an actual function has never been demonstrated. This is partly due to the nature of the previous studies on H. aspersa courtship, which have tended to be anecdotal (i.e. Tryon,

1882 i Taylor, 1900; Herzberg and Herzberg, 1962). One paper (Giusti and Lepri, 1980) does give a more quantitative description of the behaviour, but it is somewhat cursory, and does not discuss such phenornena as genital eversion stages, spermatophore transfer or the order and duration of the various courtship behav iours. Moreover, these previous studies often contradict one another, especially in their descriptions of dart shooting (Tryon, 1882; Goddard, 1962; Herzberg and

Herzberg, 1962). The questions of when and how dart shooting occurs during courtship have been 1argely resolved in a related species, Helix pomatia (Jeppesen, 1976; Lind, 1976).

Unfortunately, although H. aspersa and H. pomatia are very similar in their ecology and internaI anatomy, their mating behaviours are substantially different (Table 6,

Chapter 2). Therefore, the sexual behaviour of !L.. aspersa must be described separately; it cannot be assumed that the characteristics of courtship and dart shooting will be identical to that found in H. pomatia. 16 In nei ther species is the function of dart shooting nor the physiological control of courtship known. One difficulty in studying the courtship behaviour of either species, is that, unlike other, simpler, behaviours, it cannot be reliably elicited by the same external stimulus.

The problem of behavioural variability

"Under the most rigorously controlled conditions of pressure, tempe rature , volume and humidity the organism will do as it damn weIl pleases." (p. 26, Weiss et al., 1982). Thus states the scnewhat pessimistic Harvard Law of Animal Behaviour. It summarizes the common observation that any animal, from mollusc to mammal, will vary in its behavioural responsiveness to a stimulus, even under constant conditions. Some behaviours, such as feeding and sel', are notorious for the inability of the appropriate external stimuli to reliably induce them. Behaviours like sel' and feeding are often called 'motivated behaviours' to distinguish them from the more simply controlled reflexive behaviours. Initially, variability in an animal's behaviour, even for that observed in invertebrates, was .:lscribed to the workings of the animal's conscious intellect (i.e. r Romanes, see Boakes, 1984). Later worke:rs such as Loeb 17 and Watson advocated the view that animaIs were mere 'mindless machines' (see Kendler, 1987). This left the problem of behavioural variability in need of a mechanistic solution. In response, Craig, Lorenz, and Tolman developed concepts involving 'intervening variables' in an attempt to explain behavioural variability (see Kendler, 1987). An intervening variable 'explained' variations in behavioural responsiveness by inferring the existence of internaI processes which resulted in the variability of the behaviour. For example, food deprivation increased the intervening variable called hunger, and this increased the animal' s responsiveness towards food stimuli. Intervening variables tended to be specifie for a particular motivated behaviour, such as feeding. Adding to the behaviour-specific concept of intervening variables, Hull proposed that animaIs also possessed an internaI process called 'drive', which determined the intensity of a behavioural response. changes in the level of 'drive' resulted in fluctuations in the intensity of a behaviour. Unlike an intervening variable, Hull's general 'drive' was not specifie for any particular behaviour, but had an effect on aIl motivated behaviours (Kendler, 1987). Later, Hebb developed a concept of general 'arousal', an idea which was akin to Hull' s theory of general { 18 'drive' in that it too was non-specifie and, when activated, resulted in increased responsiveness in aIl motivated behaviours. Hebb (1955) proposed that general , arousal', which increased both the animal' s general level of aetivity as weIl as its responsiveness to environmental stimuli, was mediated, at least in mammals, by a brain structure called the reticular formation. In this way, Hebb, unlike earlier workers, gave his motivational hypothesis a plausible physiological mechanism (Kendler, 1987). At the sarne time, other workers were postulating the existence of a bewildering variety of 'drives' specifie for a partieular behaviour (Hinde, 1970). The definitions and the concepts behind the terms 'drive', , arousal " and ' motivation' beeame (and, to a large extent, still are) eonfused. The concepts lacked weIl

defined boundaries partly because the terms were never formally proposed, but had continued to change over the years as various eX'perimenters tried to come to grips with the problem of behavioural variability (Kupfermann, 1974). Without weIl defined terms and basic concepts, the study of motivation floundered. Bolles (1975), reviewing the field of motivation in the psychological literature, found examples of theories of behavioural variability that invoked the concept of 'general drive', ( some that had only behaviour-specific 'drives', and some 19 - which used neither general 'drive' nor behaviour-specific 'drives'. Halliday (1983) rather despairingly concluded that " ... there are no unifying principles or widely accepted theories in the study of motivation. There is probably less consensus about the nature of motivational rnechanisms than there is in any other area of ethology."

(p. 130).

Controversy over specifie and non-specifie 'drives'

Despi te the vagueness of most moti vational terms, i t was clear that a rigourous and testable distinction could

be made between general and behaviour-speeific 1 drive' theories (Hinde, 1970). After extensive study, it was found that different rneasures of general 'arousal', such as activity level, responsiveness to stimuli, and brain activity, did not always correlate (Hinde, 1970). This suggested that these phenomena were not aIl positively rnodulated by the same underlying rnechanism as was predicted by the general 'drive' or 'arousal' theory. This led to the collapse of Hull' s theory of general 'drive' • After reviewing the evidenee, MeFarland and Sibly (1972) concluded that general motivational constructs such as 'arousal' or 'drive' aceounted for an insignificant part of behavioural variability and that behaviours were controlled largely by mechanisms specifie 2

20 ta each motivated behaviour. After this, the concepts of general 'drive' or 'arousal' no longer played a prominent part in either psychologieal or ethological behavioural theories (Toates, 1986). Work was concentrated on motivational processes specifie for a particular behaviour, such as feeding. From this work arose a number of behaviour-speeific motivational theories of animal behavlour (i.e. see

Gallistel, 1980; Toates, 1986). Most postulate the existence of neural centres which are responsible for the control of specifie motivated behaviours. Behavioural variability is partly explained in these theories by the way in which different behaviours interact with each other (MeCleery, 1983). Neurophysiologists, however, continued to work on the retieular formation, which is still viewed as exerting a ,1 general effect on an animal's behavioural responsiveness 1 (Trulson and Jacobs, 1979). Work on eircadian rhythrns aiso stresses, and even assumes, that sorne rnechanism of general 'arousal' exists (Zucker, 1983). As animaIs maye from sleep to wakefulness, activity increases as does the responsiveness to aIl behavioural stimuli (Halliday, 1983) • This work suggested that general rnotivational mechanisms make an important contribution to behavioural variability, contrary to the conclusion reaehed by McFarland and Sibly (1972).

\ ] J 21 Despite the lack of agreement over many issues in motivation, most workers believe a real understanding of the phenomenon will occur only once the physiological substrates mediating variability are understood.

Unfortunately, studying the complex problem of

behavioural variability in any animal is a daunting

challenge, and technically impossible in most. Even the seerningly simple question: 'do general, as opposed ta specifie, motivational processes signif icantly influence behaviour?' has proven to be exceedingly difficul t to

answer in the standard animal of psychological research,

the white rat. Technical problems, including the lack of identifiable neurons, have made a neurophysiological

explanation of motivation slow in developing. Moreover,

in higher animaIs, a quest ion su ch as this is made even more complex by the ability of the animal to rnodulate its

behaviour using higher arder cognitive processes. In

higher animaIs, any stimulus carries with it some

'affective' and 'learned' camponents which may differ for

each animal and which can al ter the animal' s response

(Epstein, 1982). To try ta separate these factors, sorne

physiologists, for example those working on drinking motivation, differentiate between primary drinking, which

is due to tissue deficits, and secondary drinking, which

is due to other 'causes' such as social factors (see 22 ( McFarland and sibly, 1972). This technique has met with limited success.

A more manageable motivational model - the Mollusc

Molluscs, like mammals, show behavioural variability, but unlike mammals have both a numerically simpler and technically more accessible nervous system. Given the ubiquity of the problem of behavioural variability, it is likely that understanding a simpler system will give us insights into the possible basic mechanisms that operate in more complex systems. "If phenomena such as learning and motivation are emergent properties of aggregateb of neurons, then motivation in higher vertebrates may involve no new principles over similar phenomena in insects and moll uscs. The processes can be considered homologous entities, albeit in a more elaborate forme Thus motivational concepts are important in simple animaIs because they act as models for processes in more complex animals" (p. 140, McCleery, 1983). (McCleery, it should be noted, works on vertebrates). Moreover, al though invertebt'ates are capable of many complex learning tasks (see Mpitsos and LUkowiak, 1985) they are incapable of the type of information integration observed in the rat (Kandel, 1976). This me ans that there are fewer converging processes determining the ( 23 behaviour of invertebrates. This makes invertebrate behaviour less quixotic in its responsiveness to external stimuli. This in turn rnakes it easier to both define and measure motivational variables, 2 other distinct problems in motivational studies in vertebrates (Hinde, 1970) . For these reasons, molluscan models of motivation have proven to be very useful to neurobiologists in their search to understand the neural basis of behavioural variability (Weiss, et al., 1982). progress has been steady, but, as was the case with vertebrates, controversy has arisen over the question of the relevance of 'arousal' as a general motivational variable.

Controversy over specifie and non-specifie 'drives' in molluscs

Despite the unpopularity of the concept of a general motivational variable with behaviourists, Kupferrnann and

Weiss (1981) postulated the existence of a central arousal system in Aplysia that is capable of affecting diverse behaviours in the animal. Their concept of

'arousal' harks back to the ideas of Hebb (1955) who postulated the existence of a central • arousal' system which was excited by novel or noxious stimuli and which in turn positively modulated a wide variety of 24 behavioural responses. For example, Dier inger et al. (1978) define 'arousal' in Aplysia very similarly as " an intervening variable that infl \.lences general activity and responsiveness of the animal." (p. 20). As in vertebrates, the hypothesis of central 'arousal' in molluscs like Aplysia was partly inspired by the fact that they vary in their responsiveness to stimuli and in their level of activity over a 24 h period (Kupfermann, 1974). Moreover, autonomie functions like heart rate seem to correlate with the level of activity (behavioural state) as weIl as with responsiveness to external stimuli

such as food (Dieringer et al., 1978). Susswein (1984) found that food deprivation, which decreases the time Aplysia spend immobile, increases both food

responsiveness (Kupfermann, 1974) and the amount of time spent mating. Therefore, deficits in one motivated behaviour appear to affect the behavioural response thresholds of other motivated behaviours. Presumably

some central system is responsible for the correl~tion of th... diverse activities. And, as in vertebrates, noxious stimuli such as tail pinching or handling resul t in an increase in responsiveness to food stimuli (Kupfermann and Weiss, 1981). Therefore noxious stimuli seem to have 'general ' effects which could best be explained by the assumption that they are mediated by a { ,

25 - system that can modulate a wide variety of diverse behaviours. Weiss et al. (1982) also propose the existence of an executive arousal system to mediate behaviourally specifie aspects of 'arousal'. For Weiss et al. (1982) both a central 'arousal' system and an executive arousal system have neural correlates and both are important in the determination of behaviour. Kupfermann and Weiss (1981) compare their executive arousal system to the hypothalamus and peripheral autonomie system in vertebrates. Central 'arousal' in Aplysia has been likened to the sympathetic nervous system in vertebrates (Palovcik et al., 1982). Over the pa st 15 years, dedicated researchers have successively uncovered much of the neural basis of the executive arousal system for feeding (reviewed in weiss et al., 1982). Cells have been found that have the correct input and output functions to serve as mediators of food arousal (Teyke et al., 1990). However, the neural/hormonal components of the central 'arousal' system remain undiscovered. The only suggestion as to the physiological identity of a central 'arousal' system is Palovcik et al.'s (1982) hypothesis that central 'arousal' is mediated by serotonine Serotonin positively modulates both the level of activity as weIl as the responsiveness to food stimuli in both Aplysia and 26 P1eurobranchaea. Serotonin is thought to have a general 'arousing' effect in a wide variety of invertebrates

(1. e. Lymnaea, Tuersley and Mccrohan, 1988; 1eech,

willard, 1981; lobster, Kravitz, 1988).

This lack of concrete physiological evidence for a

central 'arousal' system has led to the proposal of alternatives. As in vertebrates, some workers re1y on specifie arousals or drives to account for behavioural

variability, and deny the existence of a central

1 arousal' system. Leonard and Lukowiak (1986) suggest

that Aplysia and Navanax (Leonard and Ltikowiak, 1984)

have separate and independent drives controlling each

behav iour. Their idea strongly resembles the

hierarchical assemblies of Baerends (1976). In their hypothesis, behaviours are defined functionally (Le.

feeding, sex,) . A behaviour, such as feeding, is

recognized as having a number of sub-units (1. e.

swallowing) that belong to 'feeding'. These sUb-uni ts rnay be involved in more than one behaviour, but each

functionally defined behaviour i5 centrally (and separately) controlled. It differs from the central

'arousal' model in that the control of and effects of different behaviours are mediated by systems distinct for each behaviour, and not via a central system. For example, the separate 'drive' hypothesis predicts that neurally active substances like serotonin or SCPb will be

j iJ 1 27 behaviourally specifie in their effects, and will not affect behaviour globally.

These di ffering predictions have been examined in the interaction of feeding and sexual behaviour in Aplys ia.

Susswein (1984) na!:> postulated that a central 'arousal' system positively modulates both feeding and sexual behaviour in Aplysia, as weIl as the autonomie responses related to the performance of the 2 behav iours. However,

Colebrook et al. (1984) have shown that the suppression of the gill wi thdrawal reflex (GWR) , which can be induced by either food satiation or copulation, is mediated by 2 different peptides, arginine vasotocin (AVT) and metenkephalin. They concluded from this that feeding and sexual behaviour exert effects on other behaviours via separate physiological mechanisms.

Therefore they reject the hypothesis of Palovcik et al.

(1982) and Weiss et al. (1982) that behaviours exert

'motivational' effects via a common, central system and instead postulate that each behaviour has its own mechanism. The concept of a central 'arousal' system has suffered other attacks as weIl. One of the major pieces of supporting evidence for a central system is the ability of nonspecific stimuli to elicit diverse behavioural effects (Kupfermann and Weiss, 1981). However, the same non-specifie stimulus, handling, that increases 28 responsiveness to food stimuli in Aplysia has no effect on feeding in either Navanax (Susswein and Bennet, 1979)

or Lymnaea (Tuersley and McCrohan, 1987). This raises the question of the concept' s general appl icabil i ty.

An atternpt at synthesis

The following study examines the variability in the sexual responsiveness of the terrestrial mollusc, Hel ix

aspersa. It tests to what extent the concept 0 f central 'arousal' is important in accounting for the observed variability. Because Helix is a terrestrial mollusc with a dissimilar life history from that of Aplysia, a marine mollusc, it also offers an opportuni ty te examine the

extent to which a central ' arousal' system rnay be influenced by an animal' s ecology.

Moreover, virtually aIl of the work concerning motivation in moll uses has deal t wi th feeding behaviour. But feeding behaviour is often viewed as a different kind of motivated behaviour from that of sexual behaviour

(Toates, 1986). Feeding behaviour is considered a

'homeostatic' behaviour in that the animal must maintain certain nutritional levels to survive. Food deprivation causes measurable physiological deficits (see Toates,

1986) • Not surprisingly the control of feeding l 29 behaviour, and hence its 'motivation', is usually modelled as a negative feedback loop (Toates, 1986; but see Hogan, 1980, for a dissenting view of horneostasis and motivation) . Sexual behaviour on the other hand, is usually considered as a typical example of a 'non horneostatic' behaviour. Sexual deprivation does not result in any physiologically measurable deficit

(Halliday, 1983). For this reason, its rnechanisms of control may differ from that of feeding, as there is no obvious physiologieal route for a negati ve feedback mechanism. This may alter the relevance of a central 'arousal' system to the control of sexual behaviour as opposed to feeding behaviour. This study begins with a description of the sexual behaviour of Helix aspersa (chapter 2). Chapter 3 explores the effects of dart shooting on courtship and sexual arousal.

Equally important, the slippery concepts of 'arousal', central 'arousal', and other motivational variables are defined and given rigourous operational definitians in the subsequent chapters, partieularly in chapter 4. As we will see in the general discussion, much of the controversy over the relevance of specif ic and non specifie 'drives' or 'arousals' in molluscs is due to a misunderstanding of what is meant by the term central

'arousal' . 30 The last three chapters explore the effects of different treatments and behaviours on sexual behaviaur as weIl as the effects of sexual behaviour on the expression of other behaviours such as ftaeding. This allows the assessment of the likelihood that different behaviaurs are madulated by a cammon system. 31

Chapter 2. Courtship and Copulation in the Terrestrial Snail, Helix aspersa 32 Abstract

Mating behaviour in Helix aspersa has three maj or components: introductory behaviour, dart shooting and

copulation. Introductory behaviour 1 which includes reciprocal tactile and oral cuntacts, lasts an average of 33.7 +/- 23.3 (s.d.) min. Dart shooting, the pushing of

a calcareous dart into the mating partner' s body 1 occurs once for each snail per mating sequence. Snails that

hit their partners during the first dart-shooting event

copulated 14.5 +/- B. 7 min after dart shooting whereas

snails that missed their partners took 40.5 +/- 37.5 min

to copula te. Dart shooting may facili tate mating by increasing behavioural synchrony. Copulation is

reciprocal and has a duration of 421. 8 +/- 56.6 min. Sperrnatophores are transferred approximately 300 min

after simul taneous intromission. There are significant

differences in the mating behaviours of Helix aspersa and Helix pomatia.

( ~'. 33

Introduction

Helix aspersa, a simultaneously reciprocal hermaphrodite, has a complicated mating behaviour that has intrigued biologists for over a century. However, previous accounts of courtship and copulation in IL..

aspersa give an ~ncomplete or poorly quantified description of the behaviour (Tryon 1882; Herzberg and

Herzberg, 1962; Giusti and Lepri 1980). This is surprising, gi ven the large amount of work that has been published on the reproductive physiology of this species (see Gomot and Deray (1987) for a review). In the present paper we attempt to provide a more complete description of the mating behaviour of H. aspersa. Such knowledge is especially desirable in view of recent

observations suggesting that the neural rnachinery responsible for mating behaviour in this species may be

amenable to experimental analysis (Chase 1986). Like its behaviour, the reproductive anatorny of IL.. aspersa is complex (Fig. 1). One of the most enigmatic structures is the dart apparatus, which consists of a

muscular sac containing a calcareous, bladed dart. The

dart is pushed into the partner during courtship 1 and is

used only once (Tompa 1982). It regenerates once every 6

days under laboratory conditions (Tompa 1982). The dart Figure 1. Reproductive anatomy of Helix aspersa (adapted from Tampa, 1982). The diagram is schematic and not drawn ta scale. Features pesterior te the spermoviduct have been omitted. A, atrium; AG, albumen gland; B, bursa; BT, bursa tract; BTC, bursa tract diverticulumi

DG, digitiform gland; OS, dart sac; F, flagellum; P, pen i s; SO, spermov iduct; VD, vas de ferens . 1

\\ AG F\\ 1 1 \ \ \ \ ( 34 is composed of a prote in matrix and aragonite, a form of calcium carbonate (Hunt, 1979). Darts used in mating are found in 11 out of 65 pulmonate families (Tompa 1980), and the dart apparatus may have evolved independently in seven major lineages (Chung, 1986). If this is so, it suggests sorne adaptive value for the dart. Its function, however, remains unclear. Its significance for the rnating of H. aspersa is investigated in the present study. Methods and Materials

Specimens of Hel ix aspersa were imported from California. After 1 week of acclimation in a colony box (30 x 32 x 36 cm), test snails were individually housed in Lucite containers (7 x 8 x 8 cm). They were fed lettuce and carrots ad libitum. A damp paper towel

lining the bottom of each container kept the snails moist. They were rnaintained on a 16L: 80 light cycle.

The temperature was kept at 200 +/- 2°C. The containers were cleaned once weekly. Each animal was numbered with na!l pol!sh or adhesive tape.

Snails were isolated from one another 3 d - :3 weeks before mating trials, with most snails (76%) being

isolated for 10 d. During the trials, 10 - 14 snails were placed together in a Lucite observation chamber (18 x 18 x 8 cm). As.::t pair began to court, they were 35 removed from the initial observation chamber and placed in a second container (18 x 18 x 8 cm). The phases of the mating behaviour were timed and recorded. During a trial , 62 +/- 21% of the animaIs eventually copulated

(n=14 trials) 1 though the last pair often started to court more than l h after the first pair. There were no significant differences in the mating times of animaIs that mated first and those that mated last. Also, the numbel. of days the snails had been isolated had no significant effect on mating times. Therefore, the data for aIl pairs of animaIs have been treated as equivalent and averaged for statistical purposes. Ten mating sequences were videotaped for a more detailed analysis. Results

The temporal pattern of the courtship behaviour is shown in Fig. 2. A prominent feature is the repeti tion of different components of the behaviour. The duration and frequency of the different components varies widely between mating pairs. Sorne components are always present

(Le. , L, LG, C, and P). However, dart shooting and biting do not occur in sorne sequences. The sequence in Fig. 2 is considered representative because i t conta ins aIl of the COMon behavioural components. Mating can be divided into three distinct phases: introductory behaviour, dart shootinq and copulation. , ,1 J 1 f 1

Figure 2. Temporal pattern of a representative mating sequence. The two lines represent a single continuous sequence, showing the behaviours of the mating snails, A and B. Each indicated behaviour continues until the next indicated behaviour. Behaviours lasting less th an 30 s have been omitted, except for biting and dart shooting. Most behaviours are expressed reciprocally, but one animaIs often leads the other by a few seconds. The temporal resolutior shown does not allow for the depiction of such discrepancies. A, apposed atria; B, biting; C, circling; CO, copulation; OS, dart shooting; L, lip-lipi LG, lip-genital atriumi P, pause. ---_._------

:~~:~ : Lr--/---t-Î t--i-+-++-Î1ï++-+il i++-+li 'j-++ÎÎ-+++lff+-+-ïl +-'i ~II L LG P LG LLa PLa P L P La A La PLa L LG L LG C l

mm 0 10 15 20 25 30 35 40 45

SNAIL A r Î if SNAIL B r 1 Il Îi Î 1ï Î 1Î PLGA PLGL ...îï LG A 1L A nilr La LP LLa IIL P liLa A cei L La A P A Laï il A LG îiA co

min 50 55 60 65 70 75 80 gO 95

- 36

Table 1 gives the average duration of each phase of the mating behaviour. Descriptions of each phase are given below. Introductory behaviour When the two receptive animaIs first make contact, each turns and orients itself he ad to he ad with its partner, touching the other with it tentacles. After the initial contact, one or both snails mOle forward and each raises is forefoot slightly until the mouths touch (lip l ip, Fig. 3A). The head moves sI ightly from side to side. Normally, after lip-lip behaviour has continued for 30 s - 3 min, the genital apparatus begins to evert. The eversion has six stages, in each of which an increasing amount of the normally internaI genital atrium is extended outside of the body wall. The stages, revised from Chung (1986), are described in Table 2 and il1ustrated in Fig. 4. During introductory behavlour, the eversion can vary from stages 0 to 5. After 30 s - 12 min in the lip-lip position, both animaIs move their heads to their partner' s right. Each moves its mouth over the other's partiaIIy everted genital atrium (lip - genital atrium, Fig. 3B). In 18 of

55 observed mating sequences, one or both animaIs bit their partner on the atrium durinq lip- genital atrium behavior. The bitten animal briefly withdrew, but

- TAlLI 1. Dunllon of the phases of millng behavlour ind thelr depcndence on the outcome of the IWO din-shoollnlJ eptsodcs

Intl'Oduclol"i Din shoounl by behaYlour 10 din tint anttnil [0 dan Dan shoot.n. by shool1ng b~ li nt ,hool1ng by ~ond ~ond antmal to Jnlmal anllnal copulalton Copulation-

Ali snadst 33 7!:13 J 15 7 ~18 7 8 1 ~6 2 .$2J 8%56.6 In=361 In:a36) (II- 34) '(11-22) Fi", dan shot 35 7:: 15 Q 1.$ 5:t8 70 7 5±5 6 .$170±,. 1 hlu .,.,mer In=111 111"'21 ) tll-21) (11-16) Fi", dan shot 32 1:t314 .w 5:t37 Sa 8 8±6 4 .$272±112.6 nusscs panneri ln = 1-'1 (11::0 1.$) 111-12) (II.') Second dan shoc 7 7±5 7 41'.6±61.J hus .,.,mer (II-II) (11-9) Second dan shot 9 4±7 1 .$26.8±57.6 l1USIeS paltJleri (11-8) (11-.)

Non VIi_.,. .,yen u _l1li111_ : SD 'm.nl ..... _ ", •• 11," ••01liliiii (oIao..IlIy Ille IIIIIIIlenlf .,.I.... ~ cI.rr-. p < 0 05 1~_-W",1111'1 'hl ''''0-1111. ", KNSUl·WllhIICSl ",,~calledl .~ __ f_ Ille .....rancc ul ,he ."""IIIUI'I potN.. 10 lhe ","!!dlllwll 01 bacII ~ t'ncl"" _ I/I.mali IlOl round ,n ln ....1 ,Ile ,w~ (IJU, IUw, :Incl"" ...... cou,... IMI 01111 nue ri ... .s..n o1wnn. coun,". ( Table 2. Stages of genital eversion.

Stage Description

o No eversion. Sexually refractory animals have an almost invisible

genital pore.

Genital pore bud. The genltal atrium begins to protrude.

2 Moderate eversion of penial lobe.

3 Full eversion of penial lobe. Penial opening visible.

4 Full genital eversion. Both penial and vaginal lobes are clearly

visible.

S Bulging of the vaginal lobe.

6 The penis is maximally everted from the penial lobe. Figure 3. Characteristic episodes in the mating behaviour of Helix aspersa. The episodes are arranged in

the order of their normal appearance 1 but the pictures are not aIl of the same pair. (A) Lip-l ip contact.

Animal on the left has a sta~e 2 eversion. (S) Lip-genital atrium contact. (C) Dart shooting. Arrow points to the base of the shot dart. The shaft of the dart has penetrated the recipient's body wall. The tissue that propelled the dart into the recipient is being retracted into the animal. (D) Apposed atria. Arrow points to the artially everted penis. (E) Copulatory posture. The snails are approximately 6 cm in body length when fully extended •

.' c j:

Figure 4. Geni tal eversion stages. (A) stage 1. The genital pore is lighter in colour than the surrounding tissue because it has started to protrude beyond the body wall. (B) stage 3. The penial lobe, the posterior lobe of the genital apparatus, is fully extended. (C) stage 4.

Both penial (posterior) and vaginal (anterior) lobes are visible. (0) stage 5. BuIging of the vaginal lobe. The animal is ~n the pre-dart shooting posture, with a contracted anterior body and a bulging posterior body. (

( c D 37 ... reemerged within 10 s and the lip - genitai atrium behaviour continued. Every mating sequence contains brief breaks in mutual contact (pause, Fig. 2). Usually during Iip - genital atrium contact, but also during lip - lip contact, the animaIs withdraw from each other for 5 - 105 s (mean 26.2

+/- 25.2 (s.d.) s n=136). They often remain motionless before re-establishing contact. Longer breaks also occur during mutual contact. During Iip - genitai atrium or lip - lip episodes, a snail often turns away from its partner and travels in a full circle or in one or more semicircles (S-shaped path) . The circles have a mean radius of 3.0 +/- 1.2 (s.d) cm (n=28) and the radii range from 2.0 to 6.0 cm. The mean duration of the circling behaviour (contact to contact) is 105.0 +/- 54.1 (s.d.) s (n=28). The stage of the genital eversion usually decreases during pause or circle behaviour and increases again when contact is re established. The temporarily abandoned partner usually remains motionless, except for moving its head from side to side, unless it too is circling. About 14% of the time the partners 'lose' each other while circling and courtship ceases (n=55). Circling behaviour may be one of the methods H. aspersa uses to realign itself while

courting. 38 ., Usually after lip - genital atrium episodes, but sometimes also after lip - lip episodes, each snail pushes its everted atrium against that of its partner (atrial apposition, Fig. 3D). The animaIs can copulate only when they are in this position. Oart shooting About 30 min after the st art of introductory behaviour, the animaIs adopt a pre-dart shooting posture (Table 1). ouring this time, the vaginal lobe bulges (stage 5 eversion). A ring of body wall around the dart sac and atrium begins to contract. The body anterior to the genital pore contracts and consequently the poste ri or portion nearest the shell begins to bulge with accumulated blood (pre-dart shooting posture, Fig. 3D). A midsagi ttal crease forms along the bottom of the foot

j ust behind the head. (A similar fold is seen during penile eversion.) The animal stops locomoting, then pushes against its mate, and the dart, a hollow calcareous lancehead with four blades (Hunt, 1979), is forced into the body of the partner. The force comes from the eversion of the papilla, a tissue attached to the base of the dart (Meisenheimer, 1912). The everting tissue pushes the dart of out of i ts sac and into the recipient (Fig. 3e). The dart carries approximately 2 mg of white mucus secreted by the digitiform glands (Chung,

1986). The recipient briefly withdraws « 10 s) and then Table 3. Outcomes of dart shooting behaviour in 61 courting snails.

Outcome ~umber of observations

Hit partner 43

Penetrated body wall 22

Dart retracted by shooter 11

Incomplete penetrance of body walll 10

Missed partner 5

Shooting behaviour but no dart fired2 2

Pre-clart shooting posture but no dart fired2 8

Neither pre-clart shooting posture nor dart fired3 3

lOarts eventually expelled

2S nails were found to have empty dart sacs

3S nails were found to have loacled dart sacs

".' Table 4. Oart locations in 90 dissected snails. Snails were taken from

the colony cage and were not engaged in mating at the time of dissection.

Location !'lumber of snails

Dart sac 71

Sursa tract 2

Sursa diverticulum

Sody cavity1 4

Underneath albumen gland1 57

In mantle cavity but not underneath albumen gland1

1Probably foreign darts

{ Figure 5. Effect of success or failure of the first dart-shooting event on the latency to copulation after dart shooting. Dotted columns (n=21) show the latencies of couples in which the dart hit the target snaili open columns (n=14) show the latencies of couples in which the dart missed the target snail. 60

en 50 CI:

' .. « . ' .. a.. 40 ., . Li. . , 0 W 30 (!) c( 1- 20 Z W U 0: 10 W a.. , . 0 10 20 30 40 50 60 70 80 90 100 NO COPULATION TIME BETWEEN FIRST DART SHOOTING AND COPULATION (min)

{ 39 both animaIs resume mating. The shooter shows full penile eversion (stage 6) simultaneously with, or shortly after, shooting the dart. A stage 6 eversion is never reached before dart shooting.

Tactile stimulation of the atrium is required to trigger the dart-shooting behaviour. Consequently, 42% of dart shootings occur during atrial apposition (n=95). In this position, each animal is exerting pressure on the

other's everted atrium. However, in 21 of 95 observations made, while the shooter was pushing against

the recipient snail, the recipient crawled over the

shooter's everted atrium, causing the dart to be delivered to the bottom of the foot.

Usually, the second animal shoots a dart within 30 min after the first animal has done so (Table 1). Each

snail exhibits dart-shooting behaviour only once per

ma.ting sequence.

Table 3 reports the fates of 48 darts shot from 61

courting snails. Most (43) penetrated the partner. ~en

snails retracted their darts after shooting them into their partners. In 7 of the 10, the retracted dart

hindered copulation by blocking the female opening. The

retracted dart was never returned intact to the dart sac.

Long after copulation, retracted darts could be found in

the bursa tract diverticulum, usually ahead of the

"" spermatophore (Table 4). The dart point was towards the 40 ,.. atrium, indicating that it was the animal's own retracted dart. The darts are probably dissolved in the diverticulum. Eight snails shcTtled no dart-shooting behaviour, even

though they did assume a pre-dart-shooting posture (Table 3) . All of these snails had empty dart sacs when dissected immediately after copulation. Three snails showed neither pre-dart-shooting posture nor dart shooting posture, despite the fact that they were found to have darts in their dart sacs when dissected immediately after copulation. Received darts accumulate in the mantle cavity below

the albumen gland. Very few darts are shot into the mantle cavi ty, therefore used darts must move into the area after entering the animal. How the darts are

transported is unknown. The darts appear to be

degenerating there. They have a reddish brown coating

which resembles in col our the substance found in the bursa, a known si te of spermatophore digestion in !::L.. pomatia (Lind, 1973). Snails that do not shoot a dart, or who miss their

partners during the first dart-shooting event, take approximately twice as long to achieve copulation after

shooting a dart compared with those that shoot

successfully (Table 1, Fig. 5). The difference is due to

the fact that animaIs that fail te receive a dart during 41 .' the first dart-shooting event take approximately 3 times as long to fire their own dart compared with those that do receive a dart from their partners (Table 1). Once both animaIs have shot darts, the time to copulation is short, whether or not the second dart successfully hi ts

its target (Table 1). Therefore, the second dart does not affect the timing of copulation. Copulation (intromission and spermatophore transfer) After shooting their darts, the animaIs evert and withdraw their penes frequently, usually after tactile stimulation of the atrium by the mating partner (Fig. 3D). The behavioural phases exhibited include lip - lip, lip - genital atrium, apposed atria, penile eversion, and

pause or circle (Fig. 2). Duration and order can vary considerably. After both animaIs have shot their darts, pauses become significantly Iess frequent. During

introductory behaviour, pauses occur at a mean interval

of 53.8 +/- 46.2 (s.d.) s (n=10 sequences), whereas after

dart shooting (within the Iast 10 minutes before

copulation), pauses occur at a me an intervalof 372.2 +/-

126 s (n=10) (p

(mean 11. 0 +/- 8.3 (s. d.) (n=10). Copulation does not occur unless intromission is simul taneous, though often the first simultaneously intromission lasts only a few 42 seconds and does not lead to spermatophore transfer. Usually only the second or third simultaneous intromission leads to successful spermatophore transfer. Approximately 30 s after simultaneous intromission, the animals become quiescent and assume a characteristic copulatory posture (Fig. 3E). The tentacles withdraw to half extension and the fore foot is slightly raised.

After 30 min 2 h, the tentacles are completely withdrawn and they remain so for the duration of copulation. Before copulation, the flagellum and the penis are empty. Two minutes after intromission, dissected animals have a spermatophore in their flagellum, but is does not yet contain sperm. The spermatophore has a foretail 2-

3 cm long, a body length of 0.6 - 0.8 cm, and a posterior ta il 7 - Il cm long and thinner that the foretail. After 2 - 6 h, the spermatophore, now containing sperm in the body segment, enters the penis.

Spermatophore transfer is usually 4.5 - 6.0 h after intromission (Table 5). Spermatophores were never observed to be transferred simultaneously, although in most cases tt.e membel"s of a pair transferred their sperma tophares w i th in 1 h of each other. The spermatorhore moves fram the penis into the partner 1 s bursa tract diverticulum. Most snails (15 of 18) have trans ferred the i r spermatophore 300 min after Table 5. Spermatophore transfer. The placement of the letters indicates the location of the body of the spermatophore at the time of dissect ion. Individual members of a single copulating pair of snalls are designated by the same letter.

Location of Spermatophore Body Time from Donor 's Donor' s Recipient's Reciplent 's start of flagellum penis bursa bursa copulation (h) diverticulum, diverticulum. body cavity mantie cavLty portion portion

Snalls dissected during copulation 0.05 a,a,b,b 1.5 c,c,d,d 2.0 e.e f,f 3.0 g g,h,h 3.5 i,i 4.0 J.J,k,k 1 i 4.5 m,n,o, p,p m,o,q,r n,q,r,s,s,t,t 5.0 u u 5.5 v v,w,w x,x 6.5 y y,z,z,A,A,B,B,C,C Snails dissected after copulation 15.0 0,0 24.2 E E,F,F 45.8 G,G ,H,H 68.0 l l 43 intromission, or about 100 min beiore copulation ends. The spermatophore moves up the diverticulum and is dissolved (Table 4). Again the speed of transport differs between animaIs (Table 5). The duration of spermatophore transfer cannot be

stated exactly. From Table 5 it appears that snails take 2 - 4 h to move the spermatophore from the penis into the partner's bursa tract diverticulum. A full spermatophore contains a yellowish paste filled with sperme The paste is no longer visible 24 h after intromission, and the spermatophore becomes a hollow shell with white mucus inside. The mucus still contains sperme After 68 h, the spermatophore is thin and transparent and has begun to dissolve. The inhibition of locomotion during copulation is not complete. The animaIs will partially withdraw into their shells if disturbed. AIso, one member of a pair of copulating snails may occasionally attempt to locomote away from its partner, but after tugging on the mutual connection, it usually returns to quiescence in less than

a minute. In one case, an animal layon top of i ts partner, making it possible for the partner to move without tugging on the interlocked penes. The partner locomoted 6 cm. Snails separated before spermatophore transfer can locomote while excreting the spermatophore. If separated after spermatophore transfer, they can l

- Table 6. Major differences between the mating behaviours of Helix aapersa and Helix pomatia.

Feature !!!. as persa H. pomatia1

Ouration of int roduc to ry 33.7 :1: 23.3 min 70 - 180 min behaviour ( range)

Frontal upright posture absent invariant part of introductory behaviour

Circle behaviour common absent

Biting common rare

Duration of shooter' s < l min typically > 8 min pause after dart shooting

Posture during copulatlon apposed atria spiral upright

Number of mutual typically < 3 typically > 5 intromissions before copulation

Contact with partner yes no required Co maintain immobili ty2

rime of spermatophore approx. 300 min app rox. 4.5 min transfer after intromission after intromission

Copulation duration 421.8 ± 56.6 min approx. 5 min

lOata for !!!. pomatia taken from Lind (1973, 1976) and personal

observa tians.

2Immobility in ~ aspersa lasts as long as copulation; in !!!. pomatia it

continues after copulation until the spermatophore tail has become

internalized. 44 locomote while internalising the tail. AnimaIs could usually be separated during copulation without apparent inj ury, contrary to the reports of Herzberg and Herzbt: rg

(1962) and Giusti and Lepri (1980).

At the end of copulation, each animal withdraws i ts penis. This withdrawal is usually not simultaneous and the pair will often remain unilaterally joined for more than 30 min. Both animals remain quiescent, however, until both penes have been withdrawn. Within a few minutes of decoupling, both snails begin to locomote, even if one or both of the spermatophore tails have not been completely internalized. The 1 3 cm of uninternalised tail is taken up within an hour.

Discussion

Because the behaviour of snails is easily observed, and the mating sequences have bizarre features, several earlier s~udies, such as those of Tryon (1882), Herzberg and Herzberg (1962), and Giusti and Lepri (1980), have described aspects of mating behaviour in Helix aspersa.

By contemporary standards, much of the older literature is woefully inadequate. For example, Tryon (1882) reported that the darts are fired at the moment of copulation, but this is clearly not true. Herzberg and

Herzberg (1962) pointed out the error, but the y failed to specify precisely when the darts are released. 45 The most detailed of the previous accounts is that of Giusti and Lepri (1980). In general, our observations agree wi th the i rs . They al so noted circ ling behav i our, biting, the tactile contacts during introductory behaviour, and the high percentage of darts that hit their target. In sorne cases, however, direct comparisons are difficult. For example, they state that copulation ranges between 2 h 10 min and 7 h 8 min in duration. As the average is not given, it is difficult to judge whether the snails that we observed have equivalent copulation tirnes. Sirnilarly, they note the independence of mating times and the receipt or nonreceipt of a dart, but because they do not differentiate between first and second dart-shooting events, a direct cornparison wi th our results is not possible. comparison with Helix pomatia

As expected, there are many points of similarity between the rnating behaviours of H. aspersa and ~ pomatia. For example, during introductory behaviour, both species utilize reci !)rocal and recurrent tactile contacts and both species exhibit the unusual dart shooting behavior. However, in several important respects, the rnating behaviour of H. aspersa differs substantially from that of H. pomatia, as observed by us and as described by Lind (1976) (Table 6). These differences are probably large enough to serve as an 46 ef fective barrier against hybridization where their ranges overlap (Kerney, 1976). The large difference in duration of copulation between

H. aspersa and H. pomatia implies that they differ in the mechanism (s) by which they transfer spermatophores. Whereas the transfer of the spermatophore body in lL.. pomatia occurs in less than 5 min (LiJld, 1976), 300 min are required for H. aspersa. This difference cannot be due to the difference in the time required for formation of the spermatophore, because in both species it is formed in less than 2 min after the beginning of copulation (Table 5; Lind, 1973). Helix aspersa remain essentially motionless while copulating. Similarly, H. pomatia remain motionless approximately 3 h after copulation while internalising the spermatophore tai!. Immobility in H. pomatia decreases the chance that the spermatophore tail will be broken during transfer (Lind, 1973). If the tail is broken, the sperm cannot escape the bursa tract to fertilize the eggs (Lind, 1973). lnunobility in fL. aspersa probably serves a similar function. Despite this likely similarity of function, the mechanism(s) responsible for the quiescence appears to differ in the two species. In H. aspersa, contact wi th the partner is essential for immobility. The inhibition in H. aspersa seems to be partly due to a feedback system in the atrium 47 walls and (or) penis that senses pressure or tension and inhibits locomotion. Copulating pairs return to quiescence when they pull on their mutual connection of interlocked penes. Separated pairs, however, will locomote freely. Therefore, the animaIs can move while

transferring spermatophores but they are prev(~nted from doing so when interlocked. This differs from inhibition of locomotion in H. pomatia, which occurs regardless of whether the other partner is present after copulation and regardless of whether a spermatophore has been transferred (Jeppesen, 1976).

Lind (1973) found that the spermatophores of ~ pomatia are digested in the bursa. In H. aspersa, the spermatophores do not enter the bursa, but enter the bursa tract diverticulum instead, as reported by Tompa (1984) . The spermatophores appear to dissolve in the

diverticulum. Thus the physiological significance of the bursa in H. aspersa rernains unknown.

Significance of dart shooting. At least six theories of dart function have appeared in the literature. (i) The dart May act like a hypodermic needle, injecting digitiform-qland mucus into the recipient snail. It has been suggested that the mucus

may act to increase the recipient 1 s 'sexual arousal' (Dorello, 1925). (ii) Goddard (1962), however, .. postulates that it is the trauma of the dart' s ------

48 ,, penetration that increases the snail's 'sexual arousal'. (iii) Jeppesen (1976) and Lind (1976) contend that dart shooting, when followed by copulation, decreases the shooter's 'arousal' or 'receptivity' resulting in mating suppression. This gi ves H. pomatia a periodic mating cycle which my be adaptive. (iv) Tompa (1980) suggests

that dart shooting may affect subsequent f~rtilization but does not primarily affect behaviour at ail. Cv) Giusti and Lepri (1980) bel ieve that dart shooting tests the 'mating readiness' of the partner. Darts discourage courtship from partners not ready to copulate saving the time and energy of the more aroused partner. (vi) It is also possible that dart shooting has bec orne anachronistic and no longer has a function in H. aspersa. The dart apparatus appears to be degenerating in sorne terrestrial snaiis such as Ochtheplaila turricula (Tompa, 1980). In H. aspersa, however, dart shooting affects the

timing of copulation (Fig. 5) , and therefore it does i f~ appear to have sorne behavioural function, contrary to Tompa (1980) and contrary to the hypothesis that it is an evolutionary rel ic. Aiso Jeppesen's (1976) and Lind's (1976) hypothesis, which posits an effect on the shooter rather than on the recipient, seems unlikely, given the behavioural triggering of dart shooting. Despite the statement of Giusti and Lepri (1980) that dart shooting can occur without any tactile stimulus, we never observed 49 dart shooting in the absence of pressure being applied to the atrium. This virtually guarantees that a shooting snail will shoot i ts dart into a tangible obj ect, and

indeed few snails miss (Table 3). This suggests that the

dart' s main effect is on the recipient, not on the

shooter. However, Lind (1976) also found that pairs of

H. pomatia that recei ve darts are less 1 ikely to copulate, which is the opposite of its effect in !::L.

aspersa. AIso, while H. pomatia show a strong withdrawal

response when receiving the dart (Lind, 1976), neither we

nor Giusti and Lepri (1980) noticed a strong aversive

response in H. aspersa. Therefore, there may be (0

species difference in dart function between H. aspersa

and H. pomatia. The data presented in this paper suggest that the main

function of dart shooting in H. aspersa is related to the

ideas of Dorello (1925) and Goddard (1962), who suggest

that dart shooting increases the sexual arousal of the

recipient snail. Thus, snails that hi t their partners

during the fir'5t dart-shooting event take less time to

reach copulation than do snails that miss their partners

(Table 1). Chung's (1986) recent experiments also

support this hypothesis. He found that injections of

digitiform gland extra ct increase genital eversion, and

he has evidence suggesting that the active agent is a . polypeptide • ," 1 50 In observinq snails, one gains the strong impression that the y copulate only after reaching some internaI threshold. A similar conviction has been expressed by Lind (1976). The data suggest that receiving a dart reduces the time needed for the recipients to reach this threshold. This hypothesis would explain why the second dart-shooting event seems irrelevant to the mating outcome (Table 1); the first animal has presumably already reached the threshold for mating, therefore the dart has no effect on it. This raises the question of why the second dart is fired at aIl, especially as animaIs can mate without any dart shooting. The answer may simply be that, given the generally reciprocal nature of the matinq behaviours, it would be difficul t to selectively suppress the second dart-shooting event. The repetitive tactile stimulation during introductory behaviour may be the animaIs' s main method of increasing the sexual arousal of i tsel f and i ts partner, wi th dart shooting being a mechanism to even out any 'arousal inequality' between the two animaIs. When no such inequality exists, the effect of the dart should be sliqht. Thus, in cases where a missed dart had 1 i ttle effect on courtship time (Fig. 5), it may be that the two snails were already approaching the threshold for dart shooting, independently of the occurrence of dart ( shooting. .... 51 In a natural environment, decreasing the time to copulation may be advantageous. Not only would the animaIs be vulnerable to predators for a shorter time, but the chance of the mating being permanently interrupted would also be reduced. Interestingly, the only instances of mating failure after the occurrence of dart shooting was when the first dart missed its intended target (Fig. 5). In these cases, the second dart shot was never fired. The animals parted before the second animal reached threshold. Peake (1978: p. 497) has pointed out t.ha\::; "considerable effort and risk ls expended or encountered by stylommatophoran molluscs during mating, and therefore it May be anticipated that features which reduce the risk or insure successful mating will have a high selective advantage. Il Dart shooting May be such a feature. 52

Chapter 3. The 'Love Dart' of the Snail, Helix aspersa, Injects a Pheromone that Decreases Courtship Duration ... _._------

53 Abstract Durinq the courtship of Helix aspersa, the snails push calcareous darts into one another. To examine the importance of this unusual action for the normal expression of their sexual behaviour, one group of Helix aspersa was prevented from exchanging darts by surgically removing their dart sacs. Another group of Helix aspersa had their digitiform glands surgically removed. These glands secrete a mucus that coats the dart. Glandless snails were capable of exchanging darts, but the darts lacked their normal mucous coating. 80th dartless and glandless pairs required more courtship time to reach copulation than did sham operated controls. Therefore, it appears that it is the mucus from the digitiform gland, not the mechanical action of the dart, which affects courtship duration. Injections of gland homogenate decreased the courtship duration of sexually receptive, dartless and glandless snails. Gland homogenate also increased the size of the recipient' s geni tal eversion, and retarded locomotion. The mucus was only effective if it entered the circulatory system of a sexually receptive snail at a specifie stage of genital eversion. The active agent in the digitiform gland mucus fulfills the requirements for the substance to be classified as a pheromone. It enters the circulatory 54 \· system of the conspecific via inoculation by the dart. 55

Introduction

The terrestrial snail, Helix aspersa, is a simultaneously reciprocal hermaphrodite with a

complicated courtship. Courtship can be divided into

three phases: introductory behaviour, dart shooting, and

copulation (Chapter 2, Chung, '87). During introductory

behaviour, the snails progressively evert their normally internaI geni tal apparatus; the extent of the eversion can be described as consisting of six successive stages (Chapter 2). At stage 0 the snail has no visible eversion, and at stage 5 the eversion is maximal. Stage

5 eversions occur just prior to dart shooting. During

dart shooting, the snails push a 1 cm long, calcareous,

mucous-coated dart (Hunt, '79) through the body wall and

into the hemocoel of their partner. The mucous coating

of the dart is secreted by the digitiform glands.

Usually, the dart, with its attached mucus, detaches and

remains inside the partner's body cavity. Because each

~:nail shoots a single dart in a mating sequence, each

courtship contains two dart shooting events (Chung, '87,

Chapter 2). The snail can regenerate a new dart in about

6 days under laboratory conditions (Tompa, '82).

Despite speculation for over a century, the effect of dart shooting on courtship behaviour remains uncertain

L 56 (see Chung, '87, and Tompa, '84, for reviews). Recently, however, chung (' 86) has found that the injection of boiled digitiform gland extract can elicit a genital eversion in H. aspersa. From this result Chung ('86) has suggested that digitiform gland mucus contains a

courtship pheromone. This study investigates the effects of unboiled digitiform gland mucus on courtship

behaviour. Specifically, i t examines the possibility that digitiform gland mucus contains a pheromone that

facilitates mating. Methods

Specimens of H. aspersa were imported from California (Marinus, Long Beach). The snails had an average weight

of approximately 8 9 (5.2 12.6 g). They were

individually housed in humid Lucite containers (7 x 8 x 8

cm) and were fed lettuce, carrots, and oyster shells ad libitum. The light cycle was 16L:8D. The temperature

was 17 +/- 2 0 C. The snails were marked with numbers for individual identification.

Demonstration of mucus transfer

To test if digitiform mucus is actually transferred from the shooting snail to i ts courting partner, eight r snails had their digitiform glands injected with carmine ! 57 red. The snails were anesthetized as described below. An incision was made in the body wall and approximately O. 05 ml of 1% carmine red was forced into the lumen at the base of each gland using a syringe and a 26 gauge needle. The glands were stained intensely red. The needle was withdrawn, and the wound was sutured. Three to ten days later, the operated snails shot darts into unoperated animaIs. The recipient snails were dissected to examine whether the red dyed mucus had entered into their body cavity.

To test how quickly substances can be carried by the snail's circulatory system, five snails were injected with 0.1 ml of 0.2% fast green dye dissolved in snail saline (Kerkut and Meech, '66). The spread of the dye was assessed by observing the snails' change in skin colour. Lesion experiments

To determine if dart shooting had any effect on subsequent courtship behaviour 1 ei ther the dart sac or the digitiform glands were removed. Before surgery, snails were injected with an anesthetic (2% MgCI2' 0.01% succinylcholine chloride, 0.005% streptomycin in distilled water; 0.1 mLI 9 snail; Chung, '85). An incision was made 3 to 5 mm behind and below the genital pore, transverse to the body axis. To remove the darts, 58 the dart sac was cut at its base and removed. To remove the glands, both digitiform glands were cut at their bases and removed. Sham operated controls had their reproductive organs manipulated by forceps. The

incisions were sutured using Ethicon 6-0 silk and a FS-3 needle. The mortality rate was less than 5% for aIl operations. Snails were allowed to recover for 2 weeks. When examined six months later, no regeneration of dart sac or digitiform glands was evident. For the courtship trials, 20 snails were removed from their isolation chambers at the beginning of the light cycle and placed together in a Luci te box (17 x 17 x 7 cm) . Dartless and glandless snails were placed in one box, and sham operated controls were placed in another. In four trials control snails and dartless or glandless snails were placed together. When two snails began introductory courtship behaviour, they were separated from the other snails and put together in a srnaller

Lucite box (14 x 7 x 7 cm). This allowed for easier observation of the behaviour, and prevented its interruption by other snails. If a sna!l copulated during a trial, it was not tested again for at least 10 days te allow time for the dart te regenerate. Each trial lasted 90 min, after which the snails were returned . to their home cages • Courtship duration and the time 1 between the two dart shoot!ng events was recorded. 59

Effects of digitiform gland mucus

To test the effect of digitiform gland mucus independent of the effect of dart penetration, homogenates of the digitiform glands were injected. Digitiforrn glands were ta ken from snails only if they had a complete dart in their dart sac, and had mucus visible in the lumen of the glands. Donor tissue was either used immediately or frozen at -70·C for subsequent use. There was no significant difference between the effects of frozen and fresh samples (Mann-Whitney, two-tailed, p»0.05) . Frozen or fresh donor tissue was ground with H. aspersa saline (Kerkut and Meech, '66) in a 1 ml hand homogenizer. The homogenate was spun for 5 min at 3000 9 in a refrigerated (4°C) centrifuge to remove solid debris. Homogenates were kept on ice and were used wjthin 3 h. As controls, homogenates of the dart sac and of the pedal gland were prepared in exactly the same manner. The pedal gland is a large mucous gland which is embedded in the foot and secretes the mucus that the snail uses during locomotion. Two doses of digitiform gland were used. The 'high'

1 dose of 4.5 mg tissue/g snail was approximately equal to f injecting the digitiform glands of one animal into !• .r another • The 'low' dose was 2 mg tisnue/snail.

, " ~, 1 60 Accordinq to Chung ('86), the amount of mucus carried by a dart is 2 mg. Therefore, 2 mg of digitiform gland (of which only a fraction would be mucus) shouid not exceed the am ou nt of mucus an animal receives during naturai dart shootinq. Before being given an injection, courting snails were separated from their partners. Once their genitai eversion level became stable (no changes in eversion level for 10 s), the snails were injected with 0.1 ml of either an homogenate or of saline using a 26 gauge needle in the posterior foot just below the shell. The persen scoring the response did not know with what substance the animal had been injected. Snails that secreted a mucous foam, or withdrew for more than 1 min in response to the inj ection, were assumed to have been inj ured and were excluded from the analysis.

Snails that have been hi t by a dart tend te slow their rate of locomotion (Chung, '86). Therefore, a 'low' dose of digitiform gland homogenate was tested to see if it would decrease locomotion. Snails were placed at one end of an inclined (15°) board covered with paper towel. Because they have a negative geotaxis, the snails crawl along a more or less straight line under these conditions. The distance an active snail travelled was determined by measuring the length of a black thread that

was superimposed on the residual mucous trail. A 61 baseline measure of locomotion was obtained during a la min trial prior to any injection. After taking the baseline measurement, snails were then randomly assigned

to one of four injection groups: 0.1 ml saline, or 2 mg of dart, digitiform or pedal gland tissue in 0.1 ml saline. The distance the animaIs crawled during a la min period immediately after the injection was compared to the distance travelled prior to the injection.

If digitiform gland mucus normally influences courtship behaviour, injections of the mucus should restore normal courtship durations to dartless and glandless courting pairs. To test this, unoperated snails which had reached genital eversion stages 2 to 5 were randomly assigned to a low dose of either digitiform gland or dart homogenate. Immediately after the injection, these snails were randomly placed wi th other unoperated snails that had shot a dart within the last 15 min. Post-dart shooting snails were used as partners because they are both consistent and persistent suitors. The time until the injected snail shot its own dart was recorded, as was the time required for the pair to achieve copulation. Snails that secreted mucous foam or withdrew into their shells after the injection were excluded from the results. 62

Results

Demonstration of mucus transfer

In aIl eight of the animaIs injected with carmine

red, both the atrium and the digitiform glands were

stained red. The dart itself remained unstained.

In cases where the shot dart had fully penetrated the

recipient's body wall (7 of 8), the recipient contained

red mucus in its body cavity. Some red mucus was still

attached to each dart. The mucus appeared to be

digitiform gland mucus because it was thick and opaque,

whereas mucus from the atrium is thin and transparent.

Therefore, these experiments demonstrate that at least

some digitiform gland mucus enters the recipient snail

during normal dart shooting.

Injections of fast green dye turned the entire snail

green, from posterior foot to tentacle tips, in less than

15 s. Therefore, it is clear that once dissolved in the

hemolymph, substances can be quickly distributed

throughout the animal. The limit to the spread of the

active agent(s) of the mucus, then, is the solubility of

the agent(s) in the hemolymph.

r l 2

\ 63

Lesion experiments

Dartless snails exhibit aIl the behaviours normally exhibited by intact animaIs, including characteristic postures before and during dart shooting (Chung, '87, Chapter 2). Therefore, for purposes of comparing the courtship durations of dartless snails and dart shooting snails, the dartless snails' 'phantom' dart shooting was scored as a dart shooting event. The effect of removing the dart sac or the digitiform glands was to skew to the right the distribution of intervals between the two dart shooting events, relative to sham operated controls (p

0.01, Kolgomorov-Smirnov test), but the latter two groups do not differ significantly from each other. <40

30 A

0 CI) .--as Q. CI 40 .-...C as 30 B E ~ 20 0 CD 10 asCI ...c 0 CD (,) ~ r' Q) l Q.- t 40 f,

30 C

0 0 10 20 30 <40 50 eo 70 80 Time between dart shootings (min) Table 1 Courtship and eopulauon durations of dart1ess. g1andless and sham

operated snails as a funetion of the stage of genital eversion of the dart

recipient during the first dart shoot lOg event Dart1ess or gland1ess snall

pairs were compared to sham operated control pairs using a 2·tailed Mann-

lJhitney test PaHs that falled to copulate were given an arbltrary tlme of 90

mln (end of observatlon period) However. the time from first dart shootlng to

copulatlon was calc'..Ilated uSlng only values from successfully matlng palrs

Starred values dlffer s:..gr,lflcantly (p

3, mating success .... as : O .. €l for dartless paus and glandless palrs (x2 , 0 l

>p'>O 05)

Eversion stage Opera ted or Tlme from first Number of

of reclpient control dart shooting to pairs that

copulation fail to

(ml.n + 5 e m ) copulate

(percent of tot.?l)

o or l dartless or glandless 85 1« (93 %)

sham opera:ed controls 12 (100%)

2 or 3 dartless or glandless 34 6 +j. 4 5* 3 (14%)

sham operated controls 25 5 +j- 2 0* o (0%)

4 or 5 dartless or glandless 14 1 +j- 2.2 o <0%)

sham operated contro1s 17 3 +j - 2 7 o (0\) Table 2 Courtshlp duration of unoperated dart recipients with genital evers10ns at stage 2 or stage 3 The di fference bet..,een the two starred values is significant (p-O 04, 2-talled Mann-lJhitney)

Dart shooting by first Dart shooting by

snall to dart shooting second snall to

by second snali copu1atlon

(lI11n + sem ) (min + sem)

dartless or ")- 3 ... /- 4 5 * 7 3 +j- 16 glandless n-21 n-18

sham operated 15 9 +/- l 9 * 9 4 +j- l 6 controls n-21 n-21 64 for aIl pairs. Moreover, snails hit by a dry dart, or by no dart at aIl, were more likely to fail to copulate (Table 1). No significant effects of receiving a dart were evident in snails that had either minimal eversions ( stages 0 and 1) or maximal eversions (stages 4 and 5) . In the former cases, mating failure was common regardless of whether darts were received, and in the latter cases, mating followed rapidly and irrespectively of the dart shooting event (Table 1). The behavior of mixed pairs, consisting of a control snail and a dartless or glandless snail, depended on whether the f irst dart shooting event was perform-ad by the control snail or by the dartless or glandless snail. If the first dart shooting sequence was

performed by the control snail, the co~rtship duration of

the pair was the same as if both snails had been controls

(p» 0.05, Mann-Whitney 2-tailed test). If the first dart shooting sequence was performed by the dartless or glandless snail and the control operated recipient had a stage 2 or stage 3 genital eversion, courtship duration

was increased by the same extent as i t was in dartless

and/or glandless pairs (p» 0.05, Mann-Whitney 2-tailed test) .

The frequen~y of mating was recorded for each group of operated snails. Dartless snails (n=18) mated an average of 7.3 +/- 1.8 (s.d.) times over the 23 courtship trials, while the average number of matings for glandless l 65 snails (n=17) was 8.0 +/- 1. 7 (s. d. ). The average number of matings for sham operated controis (n=2l) was 7.0 +/- 2.3 (s.d.). The mating frequencies of dartless and glandless snails were not significantly different from those of contraIs (p» 0.05, Krusk:\l-Wallis test). Therefore, removing the dart sac or the digitiform glands did not affect a snail' s tendency to mate.

Effects of digitiform gland mucus

Injections of digitiform gland homogenate increased the geni tai eversions of recipients that had stage 2 or stage 3 eversions prior to injection (Table 3). It had little effect on snails with small or nonexistent genital eversions. The increase in eversion stage caused by an injection of digitiform gland homogenate began L 2 +/-

0.8 (s.d.) min after the injection (n=22). The effect lasted an average of 8.1 +/- 7.5 (s.d.) min. In six animaIs, the effect lasted more than 20 min. Injections of dart sac homogenate, or of saline, had no positive effect on genital eversion at any genital eversion stage. Digitiform gland homogenate caused a significant decrease in locomotion (p

were analyzed using a G·test, wlth Williams correction, for each eversion stage and dosage level Hlgh dosage injections were effective lrrespective of the

lnitial eversion stage Ip

Number respondlng wlth an

Everslon stage n lncrease ln genital ever~lon

(percent of total)

Hlgh homogenate dose dlgltiform 0·1 2~ 4 (16 7%)

2·3 22 14 (64 0%) dart 0·1 15 0 (0% )

2·3 18 0 <0%) sallne 0·1 8 0 (0%)

2·3 12 0 (0%)

Low hornogenate dose digltlform 0·1 10 1 (10% )

2·3 10 ,. (40%) dart 0·1 10 0 (0%)

2·3 10 0 (0%) sallne 0·1 10 0 (0%)

2·3 10 0 (0%) Table 4 Effect of digiciform gland homogenace on 5n.11 locomotion Snuls

were allowed free lOCOmOC10\1 on an lnclined plane before and after inj ectlons

The data were tesced for normalicy and homogeneity of varlances before uSlng a

planned comparison test Wlth the effect of digitiform gland homog.nace compared

to the three controls The digltlform gland signlficantly decreased locomotlon

InJected macerial Change in distance travelled

(cm + /- se. m. )

digltiform gland 22 -8 1 +/- 3 1

pedal gland 11 +1 8 +/- 4 4 darc sac 7 +2 9 +/- 6 7 saline 16 +7 6 +/- 3 3 66 Injections of digitiform gland homogenate also decreased the courtship duratian of dartless and/or glandless pairs relative ta similar pairs injected with dart sac homogenate. Because dart sac homogenate had no

effect on either geni tal eversion (Table 3) or locomotion

(Table 4), i t is an appropriate control for the

nonspecif ic effects of inj ecting a tissue homogenate.

Seventeen snails at genital eversion stage 2 or 3 were injected with digitiforrn gland homogenate. These snails

achieved copulation with their post-dartshooting partners

in 31.8 ± 28.4 (s.d.) min. Seventeen similar snails

injected with dart sac homogenate required 60.2 ± 41.1

(5. d.) min to achieve copulation. The difference in courtship duration between the two treatments was significant at a prot.abili ty of p=O.052 (2-tailed Mann

Whitney test). The digitiforrn gland injections were effective only in snails that had a stage 2 or stage 3 genital eversion prior to the injection. Digitiform gland mucus had no effect on the courtship duration of snails with genital eversions of stage 4 or stage 5 (p»

0.05, 2-tailed Mann-Whitney test).

Discussion 67 Discussion

Our resul ts demonstrate that dart shooting, accompan ied by mucus transfer, decreases courtship duration. Pairs in which neither snail can shoot a dart have longer courtships than controis (Fig. 1). .l'he effective agent is not the dart itself, since snails receiving a dry dart took as long to mate as undarted snails (Fig. 1). The effective agent is the mucus from the digi tiform glands which is carried by the dart into the hemocoel of the recipient. Our observations using mucus stained with carmine red confirm that such transfer typically occurs during dart shooting. Once inside the hemocoel, the soluble components of the mucus could be quickly distributed to aIl parts of the snail, as shown by the injections of fast green. Moreover, the mucus, wi thout the dart, had significant effects on courtship behaviour. When mucus was injected into courting dartless and/or glandless snails, courtship duration was reduced relative to controls. We found that injections of digitiform gland mucus both decreased locomotion ('fable 4) and increi\sed the stage of geni tal eversion (Table 3). It is likely that both of these effects contribute to the observed decrease in court ship duration. By decreasing locomotion, dart shooting could allow for more tactile contact between 68 members of a courting pair. Tactile contact is essential for advancing courtship from one phase to the next (Chapter 2). By increasing genital eversion, the dart reduces the number of subsequent stages of genital eversion that the snail must pass through before it can reach i ts own dart shooting thresho1d (stage 5) and enter the next phase of courtship. Copulation normally occurs on1y after both snails have shot their darts (Chapter 2) • Therefore, the mucus appears to facilitate mating both

indirectly, by incr~~sing the opportuni ty for the shooter of the dart to court the recipient, and directly, by accelerating the recipient 1 s passage through introductory courtship behaviour. This increases behavioural synchrony by reducing the time that the two snails are in different phases of courtship. Because dart shooting increases behavioural synchrony, pairs that dart shoot tend to be more successful at reaching copulation (Table

1) •

It might be argued that the changes in courtship duration are due ta an artefact of the surgical procedures. This can be discounted for a number of reasons. First, it has already been established that if normal snails miss their partners during the first dart shooting event, the subsequent courtship is longer (Chapter 2). Second, dartless and glandless animaIs mated wi th the same frequency as operated controls; 69 therefore, the operation did not discourage mating. Third, copulation duration and the time between the second dart shooting and copulation in glandless and dartless pairs were indistinguishable from those of sham operated controls (Table 2). Fourth, when dartless and glandless snails had maximal eversions their latencies to dart shooting did not differ from those of sham operated controls (Table 1). Fifth, mixed control and dartless or glandless pairs had normal courtship durations as long as the first dart was shot by the control snail. Therefore glandless and dartless snails with genital eversions at stage 2 or 3 are capable of normal courtship despi te the surgery.

Al though a stimulatory role for the dart and i ts mucus has been implied in the studies of Corello ('25), Bornchen (' 67) and Chunq (' 86), studies of the mating behaviour in H. pomatia by Lind ('76) and Jeppesen ('76) and in !L.. aspersa by Chung (' 87) failed to find any stimulatory effect of dart shooting on courtship behaviour. Because the se authors did not report the eversion level of the recipients, their work cannot be directly compared wi th ours. Nevertheless, Jeppesen (' 76) appears to have found (his Table III) that dartless animaIs were about twice as likely to fail to copulate after 'phantom' dart shooting as were normal control , i• i 1 70 snails. Dart shooting may have a similar function in H. pomatia as in H. aspersa. Giusti and Lepri (' 80), working with H. aspersa, and Lind ('76) and Jeppesen ('76) working with H. pomatia, found that snails that copulate within one day of a previous mating do not perform dart shoating behaviour, and have a shorter courtship. This is not what one would predict if dart shooting is an adaptation for

f~~ilitating courtship. However, none of the cited authors reported how often the courtship af these snails resul ted in copulation. A snail in a dartless mating cannot accelerate the courtship of a slower partner (i.e. a partner at a lower stage of genital eversion) and this may decrease mating success. Secondary courtships may only be successful when both mating partners have large genital eversions simultaneously. The statistical significance of the difference in courtship duratian between snails injected with digitiform gland or dart sac homogenate was slightly less (p=O.052) than the usual criterian. One difficulty in performing this experiment was that the injections tended to disrupt the mating behaviour of the injected animal. Both our locomotion studies and our eversion studies suggest that the digitiform gland mucaus effects are behaviourally noticeable for only about 10 minutes after the injection. Thus, it is possible that the substance ...__ ...... _.------

71 had lost sorne of its effectiveness by the time the injected snail resumed courting. oespite these difficulties, the differences between the two groups is not negligible.

Chung ('86) also injected digitiform gland homogenate into H. aspersa but he àid not notice an effect on genital eversion using our 'low' dose. In his experiment, however, none of the recipient snails had pre-existing genital eversions. He discovered that boiled digitiform gland homogenate could induce genital eversion in snails without pre-existing eversions, although its effects on courtship behaviour are unknown. He found that a 5000 MW protein in the mucus was responsible for inducing the genital eversions. since Chung' s study used boiled digi tiform gland, and since this had a slightly different effect than the unboiled mucus used in our work, further experiments are necessary to determine if the same prote in is causing both effects. Chung ('86) has suggested that digitiform gland mucus induces genital eversion by increasing body wall contractions, ~{hich would cause an increase in blood pressure in the head. At the sarne time, it may aiso relax the l11uscle surrounding the geni tal pore, which would allow t'ne extrusion of the geni tal apparatus. This idea is supported by Bornchen's ('67) finding that 72 digitiform gland homogenate increases heart rate, as well as the amplitude of cardiac contraction, thereby increasing blood pressure in the head and the foot. Topical appl ication of the mucus (Chung, '86), or ingestion of the mucus (our observations), has no effect on genital eversion. Therefore, the substance is only effective when it is injected. The active agent in the digitiform gland mucus satisfies the requirements for a substance to be classified as a pheromone: it is released to the exterior

of one animal ~!\d has a specifie effect on a recei ving

animal of the sarne species (Karlson and Luscher, '59). Hewever, it differs from classical pheromones in that it dees not require a specifie external receptor to be effective. Instead, the substance is delivered directly

into the hemocoel by the dart. Karlson and Luscher (' 59) suggested that pheromones that reach the blood stream directly be considered as a separate class of compounds.

They suggested the term 'telemones 1 for such substances. We pre fer , however, that the term pheromone be used te describe the active agent in the digitiform gland mucus, as no use fuI purpose is presently served by a proliferation of terms. The unusual method of pheromone transfer (1escr ibed

here is probably not limi1.~d to H~lix, though this i5 the r only example known to us in which an inoculate has been Î

73 shown to contain a pheromone. Other animals, however, have the necessary behaviours and anatomy to transfer pheromones in this way. In some terrestrial snails, the dart is hollow and shaped like a true hypodermic needle through which gland products are inj ected directly into the mating partner (Tompa, '84). In some species of Euscopions, a genus of scorpion, the males sting their mates during courtship (Weygoldt, '77). Arnold and Houck ( '82) found that the salamander Desmognathus ochrophaeus bites its prospective mate during courtship, allowing secretions from the mental gland to enter the wound. The nudibranchs palio zosterae and palio dubia inoculate their partners with sperm during copulation (Rivest, '84), as do some flatworms (Hyman, '51). Further work may reveal that these or other animaIs also use inoculation as a method of pheromone transfer. The effects of the digitiform gland on genital aversion, as weIl as its effects on courtship behaviour, were consistently confined to animaIs with genital eversions at stages 2 and 3. Comparable resul ts have been oDtain1!d in other pheromone systems, where the recipient responds to the pheromone only in certa in seasonal or behavioural contexts (Shorey, '77). For example, the sex pheromone of the land snail Euhadra peliomphala is only effective during the snails' breeding period (Takeda and Tsuroka, '79) . The contingent 74 effectiveness of the digitiform gland implies a dependence on synergistic hormonal and/or neuronal conditions. Given that at least part of the neural machinery for the control of mating behaviour i5 known

(Chase, '86), further studies should be able to elucidate the mechanism (s) of this dependence.

-' 75

Chapter 4. Dissociation of Sexual Arousal and Sexual

Proel i vi ty in the Garden Snail, Hel ix aspersa 76 Abstract Sexual arousal (intensi ty of courtship) and sexua 1 proclivity (tendency to court) in Helix aspersa can be reliably measured using externally observable correlates. Snails with sexual proclivity are significantly more likely to turn l:owards an anesthetized conspecific after contacting it than are sexually unreceptive snails. Sexual arousal can be inferred from the stage of a snail's genital eversiol which appears only during courtship. The higher the stage of the eversion, the shorter the time required to C'omplete introductory courtship behaviour, the higher the rate of successful copulation, the fewer the number of breaks and pauses during courtship, and the longer the time a snail will spend in contact wi th an anesthetized conspecif ic. Sexual proclivity has no effect on feeding or locomotory behaviour; however, sexual arousal inhibits feeding and increases locomotor activity. Snails that were allowed daily contact wlth conspecifics required less time to complete introductory courtship behaviour relative to snails that were isolated from conspecifics for 1 week. This suggests that daily contact increases sexual arousal. A greater percent age of isolated snails exhibited courtship behaviour than did snails which had experienced daily conspacific contact. This suggests that isolation increases sexual proclivity. These 77 differences indicate that sexual areusal is not merely due te an increase in sexual proclivity. 78

Introduction

Molluscs have proven useful as model systems for the study of the biological basis of simple behaviours (Kandel, 1976). Kupfermann (1974) has argued that molluscs are also appropriate for the study of complex behavioural phenomena, such as 'motivation' (the variability of an animal's response ta a given stimulus). Biologists have been chiefly concerned with two types of behavioural variability: the fluctuation in the likelihood of a behaviour and its change in magnitude or intensity (Andrew, 1974, Kupfermann, 1974). In this paper we define the likelihood of an animal to perform a behaviour as the animal's proclivity for that behaviaur, and we define the relative intensity with which it performs a behaviour as its level of arausal for that behaviour. In mOlluscs, these two aspects of behavioural variability have been studied extensively only in the context of reeding behaviour. Food deprivation has been found ta increase bath the

1 ikel ihood of feeding behaviour (Kupfermann, 1974,

Tuersley, 1989) as weIl as its magnitude and intensity (Weiss, Koch, Koester, Rosen & Kupfermann, 1982). Therefore, both proclivity and arousal appear to co-vary 79 in feedinq, suqqesting that the physiological control of each is also heavily coordinated (Weiss et. al., 1982).

Proclivity and arousal, however, may n~t always be as closely associated in othe!' behaviours as they are in

feeding. This would sugrgest a difference in the physiological mechanisms tha:t control feed'; ng behaviour from that in other 'motiva\ted' behaviours. To test whether the degree of assoGJ.ation between arousal and proclivity does actually vary in different behaviours, we have examined both phenomena in the sexual behaviour of H. aspersa. The sexual behaviour of H.aspersa has been previously

described (Chapter 2; chung, 1987). In brief, H. aspersa are simultaneously reciprocal hermaphrodites. Courtship contains 3 basic phases: introductory behaviour, dart shooting and copulation. Introductory behaviour is characterized by repeated cr,mspecific contacts. During introductory behaviour snails progressively evert their

normally internaI genital :'!pparatus. The extent of the eversion can be unambiguously scored as 6 successive stages by its size and r.;hape (Chapter 2). Towards the end of the introductory phase, snails adopt a stereotypie shape and posture (stage 5). While in this posture, snails push against their partner and then evert their dart sac, tht'!reby thrusting a 1 cm long dart through the body wall of their mate. Each snail fires 80 .' only one dart per courtship. After dart shooting the

snail reaches stage 6 at which time i t everts i ts penis and attempts to copulate.

We first operationally define sexual proclivity and

sexual arousal in H. aspersa. We then examine the degree

of independence between sexual proclivity and sexual

arousal. The two are distinguished by their differential effects on locomotion and feeding behaviour, as weIl as

by a differential dependency on social deprivation.

General Methods

H. aspersa were imported from California (Marinus, Long Beach). The snails had an average weight of about

8.0 9 (range 5 9 13 g). Those snails that were observed copulating wera transferred to individual Lucite

containers (7 cm x 8 cm x 8 cm). These snails were

rnaintained at 190 C +/- 2 0 C and on a 16: 8 L: 0 light

cycle. The containers were kept moist and tne snails

were fed lettuce, carrots and oyster shells ad. lit,. The

cages were cleaned every day or every other day. AlI

experiments were run at the beginning of the 1 ight cycle

and continued for a maximum of 2 hours. Each snail was

individually rnarked. Snails were randomly chosen for

trials either by a coin flip or by a lottery system based

on their identification number. 81 ( A snail was used in an experiment only if it had a stable genital eversion (no change in stage for 15 s after being removed from i ts partner). The genital eversion stage often fluctuated during trials, but snails were classified by their initial genital eversion. Snails were used only once per dal'. After dart sheoting, snails were net used again for another 10 days te allow the dart to regenerate. statistical anë\lyses were carried out according to

the procedures described by Sokal and Rohl f (1981).

Experirnent 1. A Test of Sexual Proclivity

Courtship in H. aspersa begins with reciprocal tentacle touching. Contact with conspecifics results in withdrawal except during courtship. Therefore, to measure sexual proclivity, snails were tested for their response to tactile contact with an anesthetized conspecific.

Methods

Fi ve snails from the laboratory colony were anesthetized by injecting them with 0.1 mL of 2% MgCI2'

0.1% succinylcholine chloride (Chung, 1985). They were ( placed in small Lucite boxes (8 cm x 8 cm x 10 cm), where l

82 they remained motionless, with extended, but flaccid, bodies. At the begj nning of the light cycle, an isolated snail was placed so as to face the tentacles of an anesthetized snail at a distance of 1 cm. Observations were made as to whether the snail approached (positive response) or withdrew (negative response) from the anesthetized conspecific after the initial tactile contact. A snail was scored as approaching if i t continued locomoting towards the anesthetized snail until both bodies were in contact. Snails that appeared to be ambivalent (did not clearly approach or withdraw) were returned to their cages and were not given the second part of the test described below.

After the completion of the initial test with the anesthetized snail, the test snails were placed together in a Lucite 'group box' (20 cm x 20 cm x 8 cm) for 2 h. AlI courtship behaviour was recorded.

Results

Snails that turned towards an anesthetized conspecific were much more likely to subsequently engage in courtship, compared to snails that withdrew from an anesthetized conspecific (G-test with Williams correction, R<.Ol). 66.7% of positively scoring snails 83 subsequently courted (54/81 snails) while only 16.8% of negatively scoring snails did so (15/89). Therefore, this test can be used to estimate a snail' s sexual proclivity.

Experiment 2. Tests of Sexual Arousal

During courtship, snails evert their genital atrium. The following experiments test whether the stage of the genital eversion, an easily scored variable, can be

correlated with the levei of a snail 's sexual a~ousal. Sexual arousal can be inferred from the magnitude and intensity of courtship behaviour, which we measure by examining courtship duration, courtship success, and the persistence of courtship behaviour.

Methods

At the beginning of the Iight cycle, 20 randomly chosen sm4ils were removed from their isola'\:.ed chambers and were placed in a group box. Thirty to forty-five min

later a second group of 20 snails was placed in a second

group box. When the snaiis in group 1 reached a stage 6 eversion, they were allowed to court, but they were prevented from copulating by separating pairs prior to reciprocal intromission. As a snail in group 2 reached a ., •1 previously assigned stage of genital eversion, it was 84 placed in a small Lucite container with a snail at stage

6 from group 1. The snails were placed 1 cm apart w i th the tentacles facing. The duration of introductory behaviour, the number of breaks and pauses during courtship, and the number of pairs that achieved copulation were recorded. In a second t('st, 20 isolated snails were placed together in a group box. When snails reached randomly chosen, pre-assigned eversion stages they were individually placed in a small Lucite box, 1 cm away from an anesthetized conspecific. The amount of time the snails were in contact with the anesthetized conspecific was recorded. The snails were considered to be in contact if the bodies of the two snails were touching for at least 10 s. Each trial lasted 5 min. Sorne of the trials were videotaped, and a naive observer measured the contact times from the video playback. The times scored by the two observers were highly correlated (Spearman rank correlation ~(14)=.92, 2<.0001). Since no systematic bias was evident in the behavioural scoring by the first observer, the analysis of data was done using only the scores obtained by this observer.

In a third test snails were given a 5 min trial with a snail model made of rubber tubing. The amount of time snails with different stages of genital eversion spent in contact with the model was recorded. 85

Results

As genital eversion stage increased, the length of introductory behaviour during courtship significantly decreased (Fig. 1) while courtship success increased

(Fig. 2). The number of breaks and pauses during courtship also decreased significantly (Fig. 3). As geni tal eversion increased, snai ls spent an increasing amount of time in contact with an anesthetized conspecific (Fig. 4). Snails often exhibited courtship behaviour while in contact with the anesthetized partner (46 of "/2 trials). This respor.se was not completely non- specifie; snails mac!e no attempt to court a rubber model of _he same size and shape, and they spent little time in contact with it irrespect ive of their initial genital eversion (Fig. 5). In summary, snails with advallced stages of genital eversion were more intense, faster, and more persistent courters than were snails wi th lower stage eversions. A snail 's stage of genital eversion, then, can be used as an externally observable correlate of its level of sexual arousal.

( Figure 1. 'l'he duration of introductory behaviour at different stages of genital eversion. The duration of introductory behaviour decreased as the stage of genital eversion of the snail increased. The trend is significant (Kendall's Tau= -.97, R<.Ol). { 50

> 0 40 ..c 1 Q) ...... 1 .0 E • Il,) o (f) 1 ....-~ 30 • c 1 .- '-.... '0 + T oc .-c 20 • :,:; E 1 1 T 0'-'" •l ~ :J 0 10 T •.1. ! 0 0 2 3 4 5 n=4 n=17 n= 1 1 n=14 n=6 Genital eversion stage

r J

Figure 2. The percentage of successful courtships at different stages of genital eversion. The percentage of successful courtships increased as the stage of genital eversion increased (Spearman r=.83, ~<.05).

! ( 100 Q) -+-1 0 ::J 80 0.. 0 U -+-1 0 60 ..!: -+-1 (Il ~ 0 40 0...... C Q) u 20 ~ Q) Q..

o 2 3 4 5 n=10 n=g n=20 n=11 n=14 n=6 n=8 Genital eversion stage

(... -~ ------.~------

Figure 3. The nurnber of breaks and pauses in courtship at different stages of genital eversion. The number of breaks and pauses declined as genital eversion stage

increased. The trend is significant (Kendall' s Tau=-

• 97 1 g< • 01) • standard errors too small to be drawn to

scale are omitted. 5.0 c E L[) 4.0

"-Cf) "...... 1 ~ E • 0 Q) i Q) en 3.0 ~ r .Q 1 T • '+-"'-.. •l o + 2.0 ! ~ ...... Q) .Q T E 1.0 J. ::J • Z 0.0 0 2 3 4 5 n= 11 n=17 n=17 n=27 n=15 Genital eversion stage )

Figure 4. Time in contact with an anesthetized conspecific at different stages of genital eversion. As the stage of genital eversion increased, snails spent an

increasing amount of time in contact with an anesthetized conspecific. The trend is significant (Kendall's

Tau=.90, g<.Ol).

) ( 2.0

Q) "0 0 E"--" 1.5 E ~ (]) ~ en ~ 1 Q)",- 1.0 E + ..... c -+-' .- • U E 1 .....0...... - 0.5 • ... c: ... 0 • u T l •1 ! • 0.0 • 0 2 3 4 5 6 n=8 n=7 n=5 n=5 n=7 n=5 n=4 Gen ital eversion stage Figure 5. Contact time with a rubber model at different stages of geni tal eversion. The stage of a sk'lail' S genital eversion is not related to the time it spends in contact with a rubber model (Kendall's Tau=-.15, ~=.88). standard errors too small to be drawn to scale are omitted. ~, 2.0

Q) "0 0 E--- 1.5 E ..c...... , Q) .- Cf) ~ 1 Q)""-." 1.0 E + ...... , ...... , .-c • U E r ...... ,0...... " 0.5 c 0 U i T •l 0.0 • • o 2 3 4 5 6 n=8 n=7 n=5 n=5 n=7 n=5 n=4 Genital eversion stage •

86 Exper iment 3 . Differential Effects of Isolation on Sexual Arousal and Sexual proclivity

Methods Isolated snails were randomly assigned to 2 groups. On day 1 both groups were placed in separate group boxes and the number and identity of courting snails was recorded. The tirne required for each sexually aroused snail to reach stage 5 was recorded. Once a snail' s genital eversion reached stage 5, it was returned to its individual container. This prevented dart shooting, and the snails did not therefore suffer a decline in proclivity (Jeppesen, 1976, Lind, 1976). Beginning on day 2, snails in group 1 were placed in a group box at the beginning of the 1ight cycle for a maximum of 2 h every day for the next 7 days. Snails were returned to their individual containers when their genital eversions reached stages 3 or 4. The snails in group 2 remained isolated, al though they were handled daily. Snails were handled by picking them up by their shells and holding them suspended in the air for 15 s. On the eighth day, group 1 snails and group 2 snails were placed in separate group boxes. The nurnber of snails showing courtship behaviour was recorded as was the time the snails required te achieve a stage 5 genital 87 ,, eversion. As before, snails were removed from the group box when they had reached stage 5.

Results Snails that had daily exposure to conspecifics achieved a stage 5 genital eversion significantly faster than did sexually isolated snails (Table 1). However, significantly fewer snails that had daily exposure to conspecifics displayed a stage 5 eversion than did those that were isolated (Table 1). Individual snails that nad reached a stage 5 genital eversion on day 1 of the experiment, and which were

subsequently isolated, required significantly more time to reach stage 5 when they were removed from isolation

one week later (Table 2). conversely, similar snails that were given daily exposure to conspecifics required less time to reach stage 5 on the final test day (day 8) compared to day l, though this difference was of marginal

statistical significance (Table 2). oespite these differences, the time required for the same individual to

reach stage 5 on day 1 and on day 8 was significantly correlated for both the daily exposed group and the

isolated group (Spearman 1:(15)=0.63, R<.Ol for daily exposed snails, Spearman .t (18) =0.48, 12<.05 for isolated

snails) • Table l The effect of daily contact wich conspecifics or social lsolatlvn ,)n

genital eversion and courtshlp behaviour The rnean tlrnes requlred ta le h"!1 1

stage 5 genical eversion are slgnlflcancly different between the dalLv ~ p0~ed

and isolated groups (2< 02, 2·tailed :1ann-Whl.tney cesti The number of

courters ln each group lS aiso slgnificancly different (2< 05, G·test 'Nl:!>

Wllliam's correctlon)

Dally exposure to lsûlaced l week from

conspeclfics (n-78) conspeclflcs (n-70)

~jumber of courters 38 (49%) 47 (68%)

~umber reachlng 23 (30%) 34 (49%)

s cage 5

Tlme co stage 5 44 .:. +/ - 3 0 63 4 +(- 4 9

(mln +/. s e,m ) Table 2. Time dependent effects of social isolation or daily contact with conspecifics on courtship behaviour. Snails exposed to daily contact with conspecifics

achieved a stage 5 genital eversion more quickly on day 8

than on day l, though the difference was only ~arginally statistically significant (B=.07, 2-tailed wilcoxon signed ranks test). Isolated snails required si~nificantly more tirne on day 8 than on day 1 to reach geni~al eversion stage 5 (Q<.05, 2-tailed Wilcoxon signed ranks tes t) . The nurnber of courting snails did not significantly change after daily exposure, but the nurnber did increase significantly in the isolated group (Q<.03, modified G-test for repeated testing of the same individuals) .

Time (min +/- ".e.m.) Nurnber of

to reach stage 5 n courting snails n

Daily contact day 1 49.3 +/-5.6 15 30 (38.5%) 78 day 8 40.3 +/-2.6 15 38 (48.7%) 78 Social isolation day 1 47.3 +/-4.2 13 26 (37.1%) 70 day 8 63.5 +/-7.5 13 47 (67.1%) 70 88 Groups that had daily exposure to conspecifics showed no significant increase in the number of pairs that

courted from day 1 to day 8 (Table 2). Snails that were

courting on day 1 tended to continue courting every day

(11/15 snails). Lind (1976) and Jeppesen (1976) reported similar result in H. pomatia.

In summary, sexual isolation increased the time snails

needed to achieve a stage 5 eversion, suggesting that

isolation decreased sexual arousal, but it also increased

the number of snails that exhibited courtship, suggesting

that it increased proclivity in sorne of the previously

non-courting snails. Conversely, daily contact with

conspecifics had a slight positive effect on sexual

arousal, but did not significantly increase the

proclivity of non-courting snails.

Exper iment 4. DifferentiaI Effects of Digitiform Gland

Homogenate on Sexual Arousal and Sexual Proclivity

The digitiforrn glands of H. aspersa secrete the mucus

that coats the dart as it is shot. Homagenates of the se

glands can increase the genital eversion stage when

injected into snails with genital eversions at stage 2 or

3 (Chapter 3). There is also evidence that the

homogenate can decrease courtship duration times (Chapter .,. . 3) • These data suggest that the mucus can increase 89 sexual arousal. In the present experiment, the homogenate was tested for its effect on sexual proclivity.

Methods Digitiform gland homogenates of 2 different

concentrations (2 mg digitiform gland/snail and 4.5 mg digitiform gland/ g snail) were prepared as previously

described (Chapter 3). Digitiform gland homogenate was injected into randomly chosen, non-mating, previously isolated snails. Non-mating snails were those that had

shown no mating behaviour during the previous 5 trials in the group box and which had tested negative on a proclivity test immediately prior to injection. Another group of similar snails received saline injections. After the injections, the snails were given a second proclivity test. They were then collectively placed in a group box and their courtship behaviour was recorded.

Daily tests and injections were given for 8 consecutive days. The above test was repeated using snails that had tested positive on an initial proclivity test and that had exhibited courtship behaviour during at least one of the last 5 courtship trials. Results ...... _------

90 Digitiform gland homogenate injections at either dose had no effect on sexual proclivity irrespective of the initial sexual tendency of the recipient (Table 3). Therefore, al though digitiform gland homogenate appears to increase sexual arousal at the delivered doses (Chapter 3) , in the present experiment it neither increased nor decreased sexual proclivi ty.

Experiment 5. DifferentiaI Effects of Sexual Arousal and Sexual proclivity on Locomotory Behaviour

Methods Snails were removed from their individual containers and placed in a group box. During a trial a randomly chosen snail was placed on a vertical paper treadmill. The presence or absence of a geni tal eversion was noted. Snails have a negati ve geotaxis and therefore tend to crawl upwards und€r these conditions. Snails were allowed to locomote freely for 5 min. Afterwards, the distance the snail travelled was measured by lay ing a black thread over its mucous trail. In a second experiment, snails were given a proclivity test prior to the locomotery trial. These snails were net placed in a group box prior to the trial. After the trial, snails were placed in a group box and any courtship behaviour was rt!corded. Table 3. The effect of digitiform gland homogenate on sexual proclivi ty. Digitiform gland hornogenate had no effect on proclivity at either a low dose (2 mg digitiform gl,ind/ snail) or a high dose (4.5 mg digi tiform glandjg snail) (in each case, p.». 1, G-test with will iam 1 s correction). The values below are the nurnber of snails that tested positive on a proclivity test (proclivity column) and the number of snails that exhibited courtship behaviour (courtship column) (+/ s.e.m.) averaged ov·er the 8 daily trials. N=10 for aIl groups.

digitiform gland homogenate saline

low dose high dose proel ivity courtship proclivi ty courtship proclivity courtship

Non-maters

O.3±O.2 o±a O.3±O.2 O±O O.3±O.2 o±a

~Iaters

1a.O±O.o 5.8±1.3 10.0±O.O 6.5±1.1 lO.O±O.O 5.9+1.2 .,. .' .J,•

Figure 6. Percentage of snails that hi te a carrot at different stages of genital eversion. As genital eversion stage increased, the percentage of snails that took at least one bite during the trial decreased. The trend is significant (Spearman ,r=. 97, 12<.001). i 100 ..... 0 ~ 80 a~ u .....Q) .Q 60 ..... a ..c...... 40 c Q) u ~ Q) 20 0..

0 0 2 .3 4 5 6 n=.30 n=10 n=25 n=36 n=lS n=15 n=8 Gen ital eversion stage 91 Results Snails with genital eversions locomoted significantIy further (29.5 +/- 8.1 cm, n=16) than did unaroused

controis (21.7 +/- 8.7 cm, n=16, 2-tailed Mann-Whitney

test, 12< . 05) •

Snails that scored positively on the proclivity test (n=23) locomoted the same distance as those that scored negatively (n=27i 2-tailed Mann-whitney test, 12».10).

As expected, there was no difference between the

locomotion of snaiis that subsequently courted (n=19) and those that did not (n=31i 2-tailed Mann-Whitney test,

12».10, corroborating the proclivity test results.

Therefore, under conditions in which sexually aroused snails exhibited increased locomotion, snails with sexual proclivity showed neither an increase nor decrease in their locomotory behaviour.

Experiment 6. Differential Effects of Sexual Arousal and

Sexual proclivity on Feeding Behaviour

Methods

Previously isolated snails were placed in a group box.

Randomly selected snails were removed from the group box and were placed in a small box (8 cm x 8 cm x 8 cm) facing a piece of carrot 1 cm away. The initial geni tal eversion of the snail, if any, was recorded. The time 92

the snail spent in contact with the carrot was recorded. Contact was defined as touching the carrot with the head or body. Touching the carrot with only a tentacle did not count as contact. Snails in contact with the carrot

were al so observed to see if they bit the carrot ( i. e.

scraped i t wi th their radula) .

A similar trial was conducted with isolated snails after they had been given a proclivity test. After the trial, snails were placed in a group box and their

courtship behaviour was recorded. Sorne trials were videotaped, and the times were scored by an independent observer. The correlation between the two scorers was high (Spearman .r ( 2 8) =.92,

2<.01), therefore only the observations of the initial scorer were analyzed. Results Snails with a higher stage of genital eversion spent less time in contact with the carrot than did those with lower stage eversions (Kendall's Tau n(7)= -.68, 2<.03).

The proportion of snails that took at least one bite was also less when they exhibited high stage eversions than when they exhibited low stage eversions (Fig. 6). Snails that had scored positively on the proclivity test (n=16) did not differ in the time they spent in contact wi th the carrot or in the proportion of snails that took at least one bite when compared to snails that had scored 93 negatively on the same test (n=34). Similarly, there was no differenee in carrot contact time, nor in the proportion that fed, between snails that subsequently eourted and those that did not. In summary, al though sexual arousal inhibited feeding behav i our, under the same cond i tions sexua l proe livit Y had no measurable effect.

General Discussion

Table 4 sununarizes the differences between sexual arousal and sexual proclivity, based on the results of this study. Though sexual proclivity is necessary for the expression of sexual arousai , conditions which increase proelivity (such as social isolation) do not increase sexual arousal. This suggests that arousal is not simply due to an increase in the same physiological conditions that create proclivity. Fig. 7 illustrates this hypothesis as part of a model for Helix sexual behaviour. The conclusion that arousal and praclivity are distinct phenamena differs fram that which would have been suggested if similar experiments had been conducted using feeding behaviour. For example, we would expect starvatian ta increase the number of snails that would feed, but we would aiso expect starvation to have a Table 4 Summary cable of che ex.p"!:-iill .. .-:~ally determined differences bec'..Jeen

sexual proclivity and sexual arousal + lncreas.:o,·, decrease, ? marglnally

slgmflcant

ProclLvlty Arousal

Dependent: 'Janab les

feeding none

locomoclon none +

Independent vanab les

social lS0 la1:10n +

dally concacc none

digiciform gland none +

homogenate

l Figure 7. A schematic sununary of the organization of sexual behaviour in Helix aspersa. Sexual proclivity results in sexual arousal, and hence sexual behav iour, only in the presence of sexual stimuli. The box labelled 'sexual proclivity' represents the sum of aIl the neural and/ or hormonal components which contribute to an increase in the animal' s responsiveness to sexual stimuli. The box labelled 'sexual arousal' represents the neural components responsible for the execution, and the speed of execution, of sexual behaviour. The filled arrows represent exci tatory effects, while the fi lled circles represent inhibitory effects. I.Dc::oIootion

sexual Arousal digitiform glarrl sexual horocqenate stimuli

Sexual sexual ---1. Proclivity cop..ù.ation isolation

aestivation

f 94 posi tive effect on their level of food arousal. Indeed, feeding experiments done using other molluscs have found that starvation increases both the tendency to feed, usually measured as the Iatency to the f irst bite or the number of spontaneous rasps, as weIl as the intensi ty of feeding, as expressed by the frequency and intensi ty of biting (Susswein, Weiss, & Kupfermann, 1978, Weiss, et. al., 1982, Tuersley, 1989). Therefore, sexual behav iour in Helix differs from feeding behaviour in these other molluscs, and probably also in Helix, in that arousal and proclivity appear to be more independent of each other. In the pond snail, Lymnaea, however, isolation both increases the number of copulating snails and decreases their Iatency ta intromission (Van Duivenboden and Ter

Maat, 1985). Therefore, in this snail, isolation has similar effects on both proclivity and arousal. The relationship between arousal and proclivity for a particular behaviour probably depends on the selective forces present during the animal' s evolution, and not on whether the behaviour is feeding or sex. The differences between the isolated and daily contact groups in the time required to complete introductory behaviour (Table 1) could occur by at least 3 different (and not necessarily mutually exclusive) mechanisms. F irst, because exposure to cons pee if ics appeared to affect behaviour 24 hours later, contact 95 with conspecifics may have residual effects that cause the snails to be already slightly aroused when they are placed in the group box, thus decreasing the time they need to reach stage 5. The effect cannot last longer than a week, however, since isolated snails show no increase in arousal. This conclusion is also indirectly supported by the fa ct that the correlation between courtship times on day 1 and day 8 is stronger for daily exposed snails than for isolated snails. The daily exposed snails seemed to maintain their relative courtship speeds, suggesting some carry-over effect. If this were not the case, une would expect the distribution of courtship speeds amongst individuals to be more or less random each day. Of course, this would not be the case if the relative courting speed of a snail is merely " due to sorne factor intrinsic to an individual snail, such as age, but this seems unlikely because then one would expect isolated snails to be equally weIl correlated.

Other animaIs have sho~'m a similar residual effect after

sexual arousai (Toates, 1980), and the phenomenon i s

also found after food arousal (Susswein et. al., 1978). A second possible explanation for the differences in courtship times between the two groups is that the daily exposed snails show a learning, or practice, effect which

increases their speed relative to isolat~d snails.

l 96 A third possibility is that isolation inhibits sexual arousal. Lack of sensory stimuli from conspeci fics may cause a decrease in the 'excitability' of the system that controls courtship behaviour, thereby requiring a longer period of sexual stimulation before the snail can reach a stage 5 geni tal eversion.

Because snails continued to court with daily exposure to conspecifics, it does not appear that exp os ure to conspecifics decreases proclivity. However, social isolation is even more effective than daily exposure in inducing or increasing proclivity (Table 1). Because both groups are sexually deprived in that neither group had an opportunity to copulate, the only difference between the 2 groups is in the amount of conspecific contact each group receives. Conspecific contact is absent during .:::.estivation, during which sexual proclivity aiso increases (Bonnefoy-Claudet & Deray,

1984). This suggests that sensory contact with conspecifics and/or their mucus can depress the development of sexual proclivity. Sensory contact with conspecifics and/or their mucus has been shown to depress growth and egglaying in H. aspersa (Herzberg, 1965, Dan &

Bailey, 1981). It seems likely that whatever initiates sexual behaviour, a different process contrals sexual arousal. proclivity biases the snail's response to specifie 97 stimuli (Fig. 4 and Fig. 5), but it develops without sensory stimuli from conspecifics, and without any specifie motor response. Courtship invol ves rece i v ing sensory stimuli from a conspecific, as weIl as producing

the appropriate motor responses. These concomitant activities in the central nervous system could be used as eues to raise or lower the intensity of courtship behaviour. Proclivity, lacking these eues, would need to

be controlled differently (i.e. hormonally). Sexual behaviour in vertebrates is also thought to be made up of mechanistically distinct components (Beach, 1976).

One of the fundamental è.ifficulties facing any attempt to rigorously study the bioloqical basis of 'motivation' , especially sexual 'motivatior,', ls the

lack of externally observable correlates. However, sexual behaviour in H. aspersa has 2 distinct, externally verifiable components: sexual arousal (behavioural intensity) and sexual proclivity (the tendeney to mate).

Moreover, sexual behaviour in H. aspersa has the

potential to be studied using the same standard neurophysiologieal techniques that have been suceessful in the study of feeding behaviour (Chase, 1986, Balaban and Chase, in press). This makes H. aspersa a convenient animal in which to study the bioloqieal basis of sexual arousal and sexual proclivity. 98

Chapter 5. The interactions of sexual behaviour, feeding behaviour and locomotion in the behavioural hierarchy of the snail Helix aspersa 99 Summary 1. Starvation both increased the number of snails that exhibited feeding behaviour (i.e. increased feeding proclivity) and decreased the latencies of response to a

food stimulus (i. e. increased food arousal) (F ig. 3). Feeding behaviour inhibited locomotion, although starvation itself had no effect (Table 3). 2. Sexual arousal increased locomotive behaviour (Table 5), but only in the absence of a mating partner. Sexual procl i v i ty had no effect on locomotion. In general, proclivity and arousal had different effects on behaviour (summarized in Fig. 9). 3. Food deprivation did not alter the preference of sexually aroused snails for sexual stimuli over food stimul i (Fig. 7). Both starved and fed courting pairs responded to a food stimulus only during periods of low sexual arousal (Table 4), al though when not sexually aroused, starved snails usually increased the amount of time they spent in contact with a food stimulus. These resul ts suggest that sexual behaviour is dominant over feeding behaviour and that the expression of feeding behaviour is dependent upon the occurrence of low levels of sexual arousal. 4. The interaction between sexual behaviour and feeding behaviour can best be described as a time-sharing arrangement. 100 Introduction

Snails, like other animaIs, normally perform only one behaviour at a time. This 'singleness of action' requires that a snail must make a 'choice' when stimuli for two mutually incompatible behaviours are presented simultaneously. The resolution of this seemingly simple problem depends upon such less than simple concepts as 'arousal', i.e. the ability of hunger and other inte4nal states to bias behaviour. Given the actual complexity of the problem, one fruitful approach has been to study an animal that exhibits behavioural choice, but which also has a minimum number of elements involved (neurons and hormones) and a minimal behavioural repertoire. One such animal, the marine mollusc, Pleurobranchaea, has been found to make behavioural 'decisions' according to ~ behavioural hierarchy, with sorne behaviours dominant over others (Davis, et al., 1974a). Because of the relative simplicity of the nervous system of Pleurobranchaea, Davis and co-workers have been able to elucidate sorne of the physiological mechanisms underlying this hierarchy (Davis et al., 1974b, Kovac and Davis, 1980a). Like Pleurobranchaea, Helix aspersa has relatively simple nervous and endocrine systems which have been used ta study physiological mechanisms of behavioural control 101 '!,.. (Chase, 1986; Balaban and Chase, in press; Chapter 3). This paper examines behavioural choice in Helix aspersa, adding to results published previously by Everett et al.

(1982) . Everett et al. (1982) found that wi thdrawal behaviour inhibits feeding, which in turn inhibits

righting, rearing and negative geotaxic locomotion. We further investigate the relationships between sexual behaviour, locomotion, and feeding.

We also examine the effect of 'arousal' on

behavioural choice. In Pleurobranchaea, food 'arousal' alters the relative positions of sorne of the behaviours

in the hierarchy (Kovac and Davis, 1980b). In this paper

we discriminate between two different usages of the term 'arousal'. 'Arousal' has been used to describe the

intensity or vigour with which an animal performs a

response, but it has also been used to describe the

likelihood that a particular stimulus will evoke an

appropriate response (Andrew, 1974). In this paper, the

intensity of a response will be called arousal, while the

tendency of an animal to perfoI'lù a given behé\viour will

be called the proclivity for that behaviour. A similar

differentiation has already been described for sexual

behaviour in H. aspersa (Chapter 4). Sexual deprivation

increases a snail's sexual proclivity, Le. it is more

likely to initiate mating, but it has no effect on a

snail's sexual arousal, i.e. the intensity with which it 102 courts. Proclivity and arousal are here tested for their separate effects on the behavioural hierarchy. The sexual behaviour of H. aspersa has been previously described (Chapter 2, Chung, 1987). Briefly, snails are simultaneously reciprocal hermaphrodites. Courtship has three main phases: introductory behaviour, dart shooting and copulation. During introductory behaviour snails evert their normally internaI genital apparatus. Prior to dart shooting, the degree of the eversion can be reliably classified into six successive stages frOlu i ts size and shape (Chapter 2).

Materials and Methods

General Methods H. aspersa were imported from California (Marinus, Long Beach) and were placed in a colony box (30 cm x 32 cm x 35 cm). The snails had an average weight of about

8.0 g (range 5 9 - 13 g). Snails that were observed copulating in the colony box were individually numbered for subsequent identification. Fifty-six numbered snails were transferred to individual Lucite containers (7 cm x 8 cm x 8 cm) to sexually isolate them. Forty other numbered snails were randomly divided into two groups and each group was placed in one hal f of a large di v ided plastic container (group cage) (16 cm x 30 cm x 35 cm). 103 1.. The container WêS divided by aLucite sheet which had been drilled with hales (diameter 3 mm) ta al10w the passage of air, but not snails. The remainder of the nurnbered snails was 1eft in the colony box. The snails were maintained at 190 C ± 20 C on a 16:8 L:O photoperiad. The snai1s were fed lettuce, carrots and oyster shells ad lib. unless otherwise noted. AlI containers were kept moist and clean. The experiments were run at the beginning of the light cycle and continued for a maximum of 3 h. Snails were randomly chosen for trials either by a coin flip or by a lottery system based on their identification nurnber. Snails were used only once per day. Only snails with a Behavioural State Score (BSS) of 4 or 5 (see below) were used for experimental trials unless otherwise indicated. statistical analysis was done according to the

procedures described in Sokal and Rohlf (19B1) unless otherwise indicated.

Behavioural State Score (BSS) Measurements

Behavioural State Scores (BSS), derived from an animal's stereotypie postures, have been found to forro a scale of increasing levels of 'alertness', or activi ty, in Aplysia (Preston and Lee, 1973) and in Pleurobranchaea (Lee and Palovcik, 1976). Because the BSS is derived 104 from body posture, the BSS estimates snail activity without disturbing the animaIs. We have divided the behavioural state of H. aspersa into six categories from o to 5 (Table 1). To test whether the BSS correlated with any other independently measured aspect of activity, randomly chosen, sexually mature (i. e. numbered) snails were removed from the colony box and placed on a large tiled surface covered with fine sand. The BSSs were taken 5 min later, and the distance each snail locomoted over the next 5 min was recorded. The distance was measured by laying a black thread over the mucous trail and measuring the length of the thread. To test the relationship between feeding and BSS, a randomly chosen, sexually mature snail was removed from the colony box and placed behind a line drawn on a glass plate. A carrot root was attached to a force transducer 1 cm away from the snail. The carrot was suspended 5 mm below and 5 mm away from the edge of the glass plate. The BSS of the snail was recorded 10 s after placing it on the glass plate. The time required for the snail to take a bite of the carrot, as evinced by the transducer records, was recorded. Table 1. Criteria for Behavioural State Scores (B55).

State Description

o Body completely withdrawn inco shell, foot not attached to substrate

1 Body not visibly excended beyond shell, but foot attached to substrate

2 Body partially extended fram shell, tentacles unextended

3 Body partially extended fram shell, tentacles visible

4 Body fully extended from shell, movement of head and/or tentacles

5 Body fully extended fram shell, whole body movement (locomotion)

l ------

105 Assessment ot Proclivity and Arousal

Sexual proclivity was determined by a test given

immediately before a trial, as described in Chapter 4.

Briefly, the test consisted of observations as to whether

a snail turned towards or away from an anesthetized

conspecific after initial contact. Levels of sexual

arousal were inferred from the stage of the snail's

genital eversion (as defined in Chapter 2) • Higher

stages of geni tal eversion correlate wi th higher levels

of sexual arousal (Chapter 4).

To demonstrate that food deprivation could increase

both feeding proclivity and feeding arousal, individually

housed snails were starved from 0 to 5 days. They were

placed behind a line drawn across a 10 cm x 8 cm x 8 cm

Lucite box, facing a carrot root 5 cm away. The carrot

was connected to a force transducer. The average feeding

proclivity of the group was measured as the number of

snails that bit the carrot. The level of feeding arousal

was measured as the latencies of individual snalls to

come into contact wi th the carrot (as evinced by the

transducer output). Contact was defined as touching the

carrot wi th sorne part of the he ad for more than 15 s.

Touching the carrot with a tentac!e did not count as

contact. Trials were 5 min in length. since H. aspersa

is able to orient to food sources using olfactory eues .. 106 { (Farkas and Shorey, 1976), it was assumed that hungry snails would exhibit directed locomotion towards the food source, whereas fed snails would move more randomly. We

show later that starvation alone does not increase locomotion (Table 3).

The Effect of Food Deprivation on BSS

Snaj ls in one half of the divided group cage were given food ad lib. snails in the other half of the cage

were given no food for 5 days. Small Petri dishes of water were supplied to help maintain the snail's hydration since they were unable to extract water from

their usual food source. After the 5 days of food

deprivation, aIl snails were fed ad lib. for 5 days, after which the previously ad lib. fed snails became the starved group and the procedure was repeated. BSSs and

copulation rates were recorded daily for each snail. Data were collected every day at the beginning of the light cycle for 100 days.

The Effect of Food Deprivation on Locomotion

Randomly chosen fed and starved snails were taken

from the group box described above and placed on a large tiled surface covered with fine sand. A snail was 107 allowed to locomote freely for 10 min after which the distance it had travelled was measured by placing a black thread over the mucous trail and measuring the length of the thread.

In a second experiment, starved or fed snails were placed on a vertical paper treadmill. Snails have a negative geotaxis and therefore tend to crawl upwards under these conditions. Prior to a trial the paper was lightly brushed with 'v-a' juice along a distance extending 20 cm above the point at which the snail was to be placed on the paper. The distance each snail travelled was measured as described above. Both experiments were scored without knowing whether a given snail was from the starved or fed group.

The Effects of Feeding and Food Oeprivation on Sexual

Behaviour

Snails in the individual containers were randomly divided into two groups. One group was given food ad lib.; the other group was starved for 22 h prior to the trial. Pairs of either fed or starved snails were placed in small Lucite boxes (a cm x a cm x a cm). Fed or starved pairs were randomly assigned to either an empty

Lucite box or to one which contained a leaf of lettuce covering the bottom of the container. The two snails

Î 108 i... were placed 1 cm apart, facing each other's tentacles. The time required for each pair to begin introductory behaviour, and the time required for at least one member

of the pair to reach a genital eversion of stage 5 was

recorded. Snails that failed to reach stage 5 were given

an arbitrary score of 120 min, which was the length of

the trial. The number and lengths of courtship breaks (>

15 s) were observed and recorded. The amount of food

consumed during the 2 h trial was determined by weighing

the lettuce before and after the trial.

The Effect of Sexual Arousal on Feeding

Snails were taken from their individual containers

and placed in a large group box (20 cm x 20 cm x 8 cm). As a snail reached a previously assigned genital eversion

stage, it was placed behind a line drawn in a small

Lucite box, facing a carrot l cm away. The time the snail

spent in contact with the carrot was recorded.

In a second experiment, socially isolated snails were

placed in a group box. When a snail ~~ached a previously

assigned genital eversion stage, it was placed 1 cm

equidistant from an anesthetized snail and a carrot. An

anesthetized snail was assumed to act as a sexual

stimulus • Snails were anesthetized by injecting them .. t with 0.1 mL of 2% MgC12 and 0.1% succinylcholine Figure 1. The relationship between a snail's Behavioural state Score and its subsequent locomotive behaviour. As the Behavioural State Score increased snails travelled significantly farther during a 5 min trial (spearrnan 1 s r=O.99, p 0 .....,~ Q) u c: 10 f .....,0 .-(f) f 0 f 0 0 1 2 3 4- 5 n=13 n=8 n=9 n=8 n=10 n=10 8ehavioural Stote Score

1 109 chloride dissolved in H. aspersa saline (Chung, 1985). The anesthetized snails remained motionless, with

extended, but flaccid, bodies. Wllether the test snail made first contact with the anesthetized conspecific or the carrot was recorded, as was the time the snail spent

in contact wi th both obj ects. In both experirnents the

tr ials were videotaped and an observer otherwise

uninvolved in this study scored the snails.

The Effect of Sexual Arousal on Locomotion

Chapter 4 has shown that sexually aroused snails travel farther on a vertical treadmill than do unaroused

snails. In a further experiment, the treadmill was again

used, except that the snails were placed 1 cm below and

facing a snail with a stage 6 genital eversion. Snails with stage 6 genital eversions are both consistent and persistent courters (Chapter 3). The effect of a potential mating partner on the distance locomoted. was measured.

Results

Behavioural state Score (BSS) Measurements

The distélnce a snail travelled in the locomotor test was positively correlated with the snail 's BSS (Fig. 1)...... ------_ ..... _------

110

As in Pleurobranchaea (Lee and Palovcik, 1976), B5Ss also negatively correlated with the latency of contact with a food stimulus (Fig. 2). Given these results, BSS will be 'ilsed in the following experiments to estirnate snail ,activity.

Assessment of Proclivity and Arousal

Twenty-two h of food deprivation increased feeding

proclivi ty. The number of snails that responded to (i. e. bit) a food stimulus was greater in starved snails (29/30) compared to fed snails (22/30, p<0.05, G-test

llsing William 1 s correction). Food deprivation also increased feeding arousal. It increased the amount of time snails spent in contact with

the carrot dur ing the 5 min trial (Spearman 1 s r (6) = 0.97, p

The Effect of Food Deprivation on BSS

As the number of days of starvation increased, activity declined (Table 2). After 2 or 3 days of food deprivation, many snails withdrew into their shells. l Table 2. The effect of starvation on Behavioural state Score.

Behavioural State Score Nurnber of 5-day trials

Fed (n=20) 2.2 ± 0.2 20 Starved (n=20) 10 Number of days starved 1 2.3 ± 0.2 2 2.0 ± 0.1 3-5 1.4 ± 0.2 Post-starved (n=2 0) 10 Number of days after starvation 1 2.9 ± 0.2 2 2.8 ± 0.2 3-5 2.4 ± 0.2

Values are given as the rnean ± standard error. The difference in BSS between starved, fed, and post-starved snails depended significantly on the number of days starved (p

20.7; Meddis, 1984). Post-starvation snails were more active than fed snails, and this difference decreased progressively with the tirne after starvation (p

the values for fed snails (Table 2).

The Effect of Food Oeprivation on Locomotion

In the absence of food stimuli, there were no significant differences between the distances locomoted

by fed snails, 22 h starved snails, or snails recovering

from starvation (Table 3). This lack of effect is not

because 22 h of food deprivation is insufficient to create any change in the snail 's behaviour: starvation at this level causes a significant increase in food arousal

(Fig. 3). Moreover, when a food stimulus was present

('V-8' juice) , fed snails Iocomoted significantly

farther than djd 22 h starved snails (p

Mann-Whitney test). Fed snails locomoted 20.7 cm ± 9.9

cm (s.d., n=20), while starved snails locomoted 10.8 cm ±

11.5 cm (s.d., n=20). The difference between the two groups can be accounted for by the fact that

siqnificantly more starved snails (14/20) stopped to feed

on the paper than did fed snails (7/20) (p

with William's correction). If the data from aIl the snails that bit the paper at least once are excluded from r the analysis, then there was no difference in locomotion i

Figure 3. Effect of food deprivation on latency to contact a carrot. As the number of days of food deprivation increased, latency to carrot contact decreased (Non-parametric 2 x 2 factorial design, Meddis

(1984), n=10, p

Fed starved Post-starved

Distance locomoted in 5 min (cm)

17.3 ± 3.0 20.B ± 3.0 21.B ± 5.3

There v:ere no significant effects of starvation on

locomotion (2-way ANOVA, repeated measures on individual

snails, F(2,38). Since there was no difference over the

f ive days of starva tion, the mean distance locomoted (± standard error) is shown for each treatment. N=20 for aIl groups. 112 between the two groups (p>0.1, 2-tailed Mann-Whitney test) . There fore, al though food depr i vat ion increased the likelihood that snails would stop locomoting to feed, i t had no effect on the speed of locomotion.

The Effects of Feeding and Food Deprivation on Sexual Behaviour

The presence of food stimul i affected sexual behaviour even in snails that had not been food deprived. Fed snail pairs that mated on lettuce took longer to reach a stage 5 geni tal eversion (Table 4) than did those tha t ma ted on bare Luci te, al though the la tency te start courtship was unaffected. Fed snails courting in the presence of lettuce took significantly longer breaks than did fed pairs courting in a box without lettuce (Table 4) • During 39% of these breaks snails bit the lettuce (Table 4) . Therefore, snails ate lettuce during ceurtship even if they were not food deprived. In contrast to fed ceurting snails, starved snail pairs required more time ta begin introductory courtship behaviour when they performed their courtship on a bed of lettuce as opposed to in an empty Luci te box (Table: 4). Prier to courtship, 100% (24/24) of the starved snails taak bites from the lettuce. Further, unlike fed caurting snails, starved snails required the same length Table 4. The effect of food stimuli on courtship behaviour

Fed Snails 22h Starved Snails Courtship Parameter No Lettuce Lettuce No Lettuce Lettuce

Latency to 17 4 ± 2 5 22 3 ± 3 3 20.5±4.Sa 328±39d tiourtsl'lip (min) n-24 Ume to 35 (27-43)b 59 (40-82)b 50 (43 -61) 44(32-49) stage 5 (min) n-24 Number failing 3 6 5 6 to reach stage 5 n-24 Percentage that 39% (22/56) 3S\ (20/52) eat during breaks n-10 Number of breaks 5 6 ± 0 S 70 ± 0.7 6.5 ± 0 6 6 5 ± 0 6 > 15 s n-lO Length of breaks 5 6 ± 2 Sb 2 5 ± 2 1 3 2 ± J. ) (min) n-10

a The indicated values are significantly different, p

snails that failed to reach stage 5 were given an arbitrary score of 120 mLn (length of trial)

1. Table 5. The effect of sexual arousal on locomotion.

Distance travelled (cm ± sem)

Alone wi th mating partner

Sexually aroused 29.5 ± 2.0 19.6 ± 2.3

Sexualll unaroused 22.7 ± 4.9 19.9 ± 3.0

Snails were given 5 min to locomote. The depression of

locomotion by the presence of a mating partner is dependent upon whether the locomoting snail is sexually aroused (2 -way ANOVA repeated measures, F ( 1 , 15) p

N=16 for aIl groups. Figure 4. The effect of starvation on the frequency of copulation. starved snai1s copu1ated 1ess frequently than fed snails, while post-starved snails copu1ated more frequently than fed snails (p

~ 0 0.08 "C 0.07 0 "c CI) 0.06 CI) c "0 0.05 :;:; 0 0.04 a.:J 0 0.03 ...0 0 0.02 ~ Q) J:l 0.01 E :J z 0.00 Starved Fed Post starved 113 of time to reach dart shooting threshold (stage 5) irrespect ive of the presence or absence of lettuce (Table 4). Therefore, food stimuli affect rnating behaviour, but the effect depends on the length of food deprivation. with little or no food deprivation, the presence of food stimuli increases the length of breaks during courtship, and deprivations of 22 h inhibit. the initiation of courtship behaviour. Snail pairs that did not mate ate significantly more lettuce (0.71 9 ± 0.33 g, n=B pairs) than did mating pairs (0.21 9 ± O. 13 9 1 n=10 pairs) during the trial whether fed or starved for 22 h (p

2 tailed t-test). Therefore, the expression of sexual behaviour coincides with reduced feeding in both starved and fed snaiis.

Starved snails copulated significantly less frequently than did fed controis (Fig. 4). However, snails fed after the 5 day starvation period copulated significantly more frequentIy than did fed snails (Fig.

4) • starvation, therefore appears to decrease proclivity, but, as with the level of general activity

(Table 2), there was a rebound effect such that levels increased to above baseline once food was again available. However, proclivity tests performed on fed snails and snails starved for 22 h showed no difference in the number of snails that exhibited a positive response (p>O. l, G-test, n=20 for both groups). The 114 number of snails that courted when starved and fed snails were placed together in a group box after the proclivity test was aiso no different for starved and fed sna ils (p>O.I, G-test, n=20 for both groups). The Effect of Sexual Arousal on Feeding As sexual arousal increased, the food stimulus was

Iess able to elicit a response (Fig. 5i Chapter 4). AIso, the nurnber of snails that chose (i.e. turned towards) an anesthetized conspecific, rather than a carrot, increased in a two choice test (Fig. 6). Further, as sexual arousal increased, snails spent an increasing amount of time in contact with the anesthetized conspecific (Fig. 7). These three results (Figs. 5-7) were not due to mating snails being food satiated, and thus having reduced feeding proclivities, because those snails that courted ate no more carrot than did those snails that did not court (0.98 g ± 0.52 g, for courting snails, n=14 i 0.78 ± 0.54 g, for non-courting snails n=12; p>O.l, 2-tailed t-test). Food deprivation did not alter the percentage of sexually aroused snails that chose an anesthetized snail instead of a caL'rot (Fig. 6). Moreover, 22 h of food deprivation did not change the amount of tirne a snail spent in contact with an anesthetized partner (Fig. 7). However, food deprivation did increase the amount of time sexually aroused snails spent in contact with the carrot Figure 5. The relationship between the stage of a snail's genital eversion and the length of time a snail spent in contact with a carrot during a 5 min trial. As genital eversion stage increased, snails spent significantly less time in contact with the carrot

(Spearman's r=-O.94, p

+J 0 5 L- L- 0 0 .l:. 4 ~ I ~,...... +J c: 0'- .3 o E +Jc '-' 0 0 2 .-c f Q) f E 1 i= f 0 0 1 2 .3 4 5 n-27 n-25 n-18 n-18 n-16 n-12 Genital eversion stage

(, Figure 6. The effect of starvation on a snail' s choice between a carrot and an anesthetized conspecific. sexually aroused, starved snails choose an anesthetized snail over a carrot as often as did sexually aroused, fed snails (p>O.l, G-test). Both starved and fed snails chose an anesthetized snail more often as their genital eversion stages increased (Spearman 1 s rank r (fed) =0.99, p

Il' G) 100 U)o [Xl starved snoils .c't-0._ 0'0 !!8 fed snails ~ G) 80 .cU)oa. ~c (1)0 ::: 0 60 o~ cG) (l)N -.-:..:; o G) 40 Q).c c,.~ CG) +.Ic ~ 0 20 oc ~o ~ n 13 8 10 12 10 8 8 10 10 10 10 10 o 1 2 3 4 5 Genital Eversion Stage Figure 7. The effect of starvation on the tirne snails spent with an anesthetized conspecific during a two cheice test. All snails spent more tirne with an anesthetized conspecific as their genital eversion stages increased (p

There was no significant difference between starved and fed snails (F(1,5)). Error bars denote ± standard error of the mean. The numbers next te each data point denote sample size. (

5 0-0 storved snoils c~ o,E e-e fed snoils .cE 4 ~ '-' ~::: ..,0 o c 3 o (1) .., "'0 C Q) o,~ 0+1 ., Q) L C~ .- +1 (1) Q) Q) c E 0 1 i=

0 13 0 , 2 3 4 5 Genital Eversion Stage Figure 8. The effect of starvation on the time snails spent in contact with a carrot during a two choice test.

AlI snails spent less time L· contact with the carrot as their gf.~ni tal eversions increased (p

Error bars denote ± standard error of the rnean. The numbers above each data point denote the sample size. 13 5 +' 0-0 starved snails 0 ~ ~ • -. fed snails c () 4 ..c 10 .-+' ~~3 13 .+Je ().- .BE g'-' 2 0 .-c: Q) 1 E .- 10 t- • 10 0 --. 0 1 2 3 4 5 Genital Eversion Stage 115 during the same two choice-paradigm (Fig. 8). To summarize, both the food deprived, and the fed, sexually aroused snails spent the same amount of time responding to a sexual stimulus, but food deprived snails spent more time in contact with the carrot when not in contact with the sexual stimulus. In contrast to sexual arousal, sexual proclivity had no effec. on whether snails turned first towards an anesthetized snail or a carrot. Snails that tested positively for sexual proclivity were equally likely to choose an anesthetized conspecific over a carrot (10/40) as were snails that scored negatively on the proclivity test (6/22, G-test, p>O.l).

The Effect of Sexual Arousal on Locomotion

Sexual arousal increased the distance snails travelled on a paper treadmill (Table 5; Chapter 4). The snails tended to travel along circular paths, which is cornrnon during courtship when rnating pairs bec orne separated (Chapter 2). However, when a snail at genital eversion stage 6 (post dart shooting snail) was placed 1 cm away from the sexually aroused snail whose locomotion was being tested, the sexually aroused snail locomoted significantly less (Table 5). The presence of snails at genital eversion stage 6 had no effect on the locomotion , 116 .. of unaroused snails (Table 5). Therefore, sexually

aroused snails l~comote faster only if they are without a partner.

Discussion

Figure 9 summarizes the resul ts of this study by illustrating the interactions between locomotion, sexual behaviour, and feeding behaviour in H. aspersa. These behaviours do not form a rigid, inflexible hierarchy, but rather the position of each behaviour is dependent upon

th~ behavioural context in which it is expressed and upon the relative levels of 'arousal' for each of the behaviours. Sexual behaviour dominates feeding behaviour under most conditions, with feeding only occurring during

periods of low sexual arousal. The nature 0 f the relationship between sex and feeding closely resembles a time-sharing arrangement (Brown and McFarland, 1979). The evidence for time-sharing, which does not seem to have been previously demonstrated in an invertebrate, is summarized below. Sexual stimuli were chosen over food stimuli by a sexually aroused snail even if that snail was food deprived (Fig. 7). Moreover, starved, sexually aroused snails spent as much time with a sexual stimulus r (an anesthetized snail) as did fed, sexually aroused Figure 9. Schematic summary of the relationships between

locomotion, sexual behaviour and feeding ~ehaviour in the snail, Helix aspersa. Unshaded arrow denotes an excitatory effect; shaded arrows denote an inhibitory effect; half-shaded arrows denote a partial inhibitory effect; open arrows signify that proclivity will produce arousal when the appropriate stimuli are present. The boxes for sexual arousal or feeding arousal signify that the behaviour can be performed at different levels of intensity. Ca) signifies excitatory effect only when there is no sexual partner; (b) signifies inhibition is incomplete during high levels of feeding proclivity or feeding arousal...... Sexual Sexual Proclivity L ~ " Arcosal

1- ~ b~, ~ ...... Feed.l.rq F~ - Proclivity " Aroosal

~ I.cXXIIDtioo ~ ~a ~------~------

117 controls. These resul ts demonstrate that when 9 i ven a choice, sexually aroused snails will chose a sexual stimulus over a food stimulus. When not in contact with the anesthetized snail, however, starved, sexually aroused snails spent significantly more time with the carrot than did fed, sexually aroused controis (Fig. 7 and 8). Therefore sexual behaviour only partially suppresses feeding. Fed, sexually aroused snails also responded to a food stimulus, and the amount of time they spent in contact with the carrot increased as sexual arousal decleased (Fig 5). This pattern was aiso seen during courtship behaviour. Eating only occurred before courtship, or during breaks in courtship in both fed and starved mating pairs. During such periods, sexual :irousal is relatively low (Chapter 2). ThE::rl5'fore, in both starved and unstarved pairs, eating only occurs during courtship sequences of low sexual arousal. This is similar to rats, where starved male rats aiso eat only

during periods of low sexuai arousai (Brown and

McFarland, 1979). During periods of low sexual arousal, elevated Ievels of feeding proclivi ty can inhibit sexual behaviour, reversing the usual relationship between the two behaviours. Fed snails take longer courtship breaks in

the presence of food stimuli, and in a starved snail, food stimuli can delay the initiation of courtship 118 behaviour (Table 4). This is also seen in the rat where starva tion can increase the latency to courtship initiation (Brown and McFarland, 1979). Nevertheless, starved snails mating either on Iettuce or in empty containers have the same number of breaks during courtship as do fed mating pairs. It does not, therefore, appear that a strong feeding proclivity can increase the periods of low sexual arousal that occur during mating. In other words, the length of time a snail spends in contact with a food stimulus during courtship is not determined by the length of food deprivation, but by the level of sexual arousal. The effects of food stimuli on the courtship of snails at different levels of food deprivation are complex (Table 4). The snails starved for 22 h responded inunediately to the lettuce, unI ike fed controls. This would suggest that a high proclivity for feeding can delay the expression of a proclivity for sex. It is possible that once the starved snails had eaten, the snails' feeding proclivity and feeding arousal declined, and the food stimulus was no longer able to inhibit sexual behaviour. The fed snails were more influenced by the food stimulus (Table 4), perhaps because they began to court before they ate and they still had a relatively high proclivity for feeding. They therefore had a tendency to take longer breaks during mating in the 119 presence of 1ettuce. This suggests that in sorne circumstances, sexual proclivi ty can inhibit the expression of a proclivity for feeding.

Feeding proclivity appears to inhibit sexual proclivity, because food deprived snails showed a decline in copulation frequency (Fig. 4). However, our test for sexual proclivity fai1ed to detect any effect of food deprivation. This apparently contradictory result can be explained by the fact that only active snails were tested for proclivity, but since starvation decreases the number of active snails, it also decreases the number of snails available for matirg (Table 2). This could explain the dec1 ine in the copulation frequency. Therefore i t appears that a procl i vit y for feeding does not affect sexual proclivity per se, but that feeding proclivity reduces the opportuni ties for mating by lessening the number of active snails.

The increases in rnating that occur during the starvation recovery period may be caused by more than

Just increased activity. Inactive snails are also sexually isolated snails, and sexual deprivation has been shown te enhance sexual proclivity (Chapter 4) . Everett et al. (1982) constructed a partial behavioural hierarchy fùr H. aspersa and found that feeding in H. aspersa has a less dominant position than in Pleurobranchaea. They speculated that this was due to 120 a difference in food strategies (carnivore vs. herbivore; patchy vs. relatively continuous food distributions). They predicted that mating would be dominant to feeding

i n :.!H..:._~a:...::s:...lp::<..e:::::...o:r:...!s::.;a:::.., contrary to the situation in Pleurobranchaea. This prediction is supported by our results. The same domination of mating over feeding is found in another herbivorous mollusc, the marine opisthobranch, Ercolamia nigra (Jensen, 1987). However, in Aplysia, which is also a herbi vorous marine opisthobranch, the interaction between mating and feeding is less easily characterized. In Aplysia, mating and feeding can occur at the sarne time (susswein et. al.,

1983) . This suggests that the organization of behaviour will not necessarily be the same even for related animaIs, but will depend upon the details of the life history of the animal. This issue has been further, discussed by Everett et al. (1982), Kovac and Davis

(1980), and McCleery (1983).

Although proclivities for either sex or feeding affect behavioural choice, the y do not have the same effect on the behavioural hierarchy as the actual expression of the behaviour. For example, sexual proclivi ty has no effect on locomotion, but sexual arousal can increase locomotion (Table 5). Sexual proclivi ty, unI ike sexual arousal, does not affect the amount of time a snail spends in contact wi th a carrot. 121 Increased feeding proclivity has no effect on locomotion

(Table 3). In these c~ses, changes to the hierarchy are dependent on the rec! Lpt of the appropriate sensory stimuli and/or the initiation of the motor program for the behaviour. Pleurobranchaea exhibits a similar phenomenon; the inhibition of withdrawal by feeding is dependent on the activation of feeding behaviour. starvation alone has no effect (Davis et al., 1977).

Given the differences in the effectiveness of arousal and proclivity in altering an animal's behavioural hierarchy, the two aspects of 'arousal' should be kept expl ici tly separate in any behav ioural hierarchy. The mechanisms that bias a snail' s response to a given stimulus (proclivity) may be very different from the mechanisms by which an aroused snail suppresses competing behaviours once the motor program for a behaviour has been initiated. 122

Chapter 6. 1 Central Arousal t and Sexual Responsiveness in the snail, Hel ix aspersa 123 Abstract

In molluscs, a 'central arousal' system is thought to positively modulate both an anirnal's level of activity and its behavioural responsiveness. This hypothesis is exarnined in Helix aspersa by testing the relationships between acti vi ty 1 feeding 1 and sexual behav iour.

Activity, feeding and mating exhibi t parallel dal1y rhythms. Snail sare most active, and eat and mate most frequently 1 during scotophase and the first 3 h of photophase. Handling and inject ions 0 f serotonin (5 x

10-7 rn/kg body wt) increase general act i vi ty. Inducing activity dur i ng l a te photophase increases food consumption, but it does not induce sexual activity.

Moreover, serotonin and handl ing have no effect on sexual arousal; treated snails show no increase in genital eversion stage and require the same length of courtship as controls. If snails are sexually deprived, increasing activity in late photophase does increase copulation frequency, but, snails do not copulate more frequently than do unhandled snails tested during early photophase.

These res ul ts suggest that 'central arousal', as estimated by the snail's level of activity, has a permissive or . gating' effect on sexual behaviour, but does not directly affect sexual proclivity or sexual arousal. Pelix appears to differ from Aplys ia in the effect of 'central arousal' on sexual behaviour. This 124 difference between the two species may be related to the differences in their reproductive strategies. 125

Introduction

A snail, like other animaIs, varies in its responsiveness to external stirnul i. Often this variability follows a predictable pattern. For example, in l'.plysia, responsiveness to food stirnul i waxes and wanes in parallel with the daily activity rhythm

(Kupfermann, 1974). An underlying variable called

. arousal' has been proposed to account for this positive correlation between an anirnal's level of activity and its

1evel of behavioural responsiveness (Andrew, 1974;

Dieringer, Koester, & Weiss, 1978). Palovcik, Basberg and Ram (1982) have postulated that a 1 central arousal' system drives increases in bath behavioural responsi veness and behavioural state (the animal' s level of acti vi ty) in Aplysia. This idea is supported by the f inding that serotonin can increase both behav ioural state and the responsi veness to food stimul i, and palovcik et al. (1982) have suggested that 'central arousal' rnay be rnediated by serotonin. In order to explore the extent to which 'central arousal' is a rneaningful variable in Helix, we have examined whether increases in , central arousal' are accornpanied by increases in behav ioural responsiveness to food and sexual stirnul i . Helix aspersa exhibits a circadian 126 activity rhythm (Bailey, 1981), and therefore we have tested the concordance of the dai1y patterns for feeding,

sexual behaviaur and acti v i ty.

Susswein (1984) has hypothesized that, in Aplysia,

not only feeding, but also sexual behaviour is positively

modulated by a 'central arousal' system. (Susswein

(1984) uses the term . common arousal', but this appears

ta be equivalent to 'central arousal'). However,

whether or not a single system can positively modulate

bath feeding and sexual behaviour is likely to depend on

the feeding and reproductive strategies of the species.

Aplysia exhibits a reproductive strategy in which it

lives through only one reproductive seasnn (Kandel, 1979)

and spends 25% of its time engaged in mating activity

(Susswein, Gev, Feldman & Markovich, 1983). tIeli}(, on the other hand, typically lives through more th an one

reproductive season and mates infrequently (potts, 1975;

Moulin, 1980). Tuersley (1989) has argued that an animal's ecology shapes the way in which its behaviour is modulated. This suggests that the neural and hormonal mechanisms controlling an animal t s mating behaviour are adaptations to i ts reproductive strategy (Crews & Moore,

1986). For this reason, it is probable that the relationship between 'central arousal', feeding and sexual behaviour will differ between Helix and Aplysia. 127

A comparison between Helix and Aplysia is desirable because Helix, like Aplysia, has proven amenable to neurophysiological analysis (Chase, 1986; Balaban & Chase, in press). The opportunity therefore exists to examine, and to compare, the physiological mechanisms responsible for modulating fee:ding and sexual behaviour in animaIs with very different life histories. To execute this test, handling of the animaIs and injections of serotonin will be used to increase behavioural state in Helix. If behavioural state, feeding and sexual behaviour are aIl positively modulated by a . central arousal' system, then feeding and sexual behaviour should increase in parallel with increases in behavioural state.

Sexual behaviour in Helix: aspersa has been previously described (Chung, 1987; Chapter 2). It consists of 3 main sequences: introductory behaviour, dart shooting and copula"...i '.>n. Dur ing introductory behaviour snails evert their normally internaI genital apparatus. The degree of the eversion can be categorized as six different stages, and these stages have been found to correlate with the level of sexual arousal (Chapter 4). During dart shooting, a snail pushes a calcareous dart into its partner. The dart has a mucous coating which is able to increase sexual arousal (Chapter 3). Increases in sexual arousal can be estimated not on1y from the stage of the 128 genital eversion, but aiso by the time a snail requires to reach stage 5 (dart shooting threshol.d) (Chapter 4).

General methods

H. aspersa were imported from California (Marinus,

Long Beach). The snails had an average weight of 9 9

(range 6 9 - 11 g). Those snails that were observed copulating were transferred to one of three types of containers: individual Lucite containers (7 cm x 8 cm x 8 cm), group containers (18 cm x 18 cm x 8 cm; 10 snaiis per box) or a colony container (35 cm x 30 cm x 30 cm; 40 snails per box). AlI snails were individually marked. Colony snails were maintained at room temperature (24 0 C

± 2 0 C), aIl other snails were maintained at 190 C ± ~o.

The light cycle was 16:8 L: D. The containers were kept moist, and the snails were fed lettuce, carrot roots and oyster shelis ad lib. ThE' cages were cleaned and the snails were fed every day or every other day at the start of photophase. AlI experiments were run at the beginning of photophase unless otherwise noted, and they were continued for a maximum of 2 h. Snai~~s we~e used anly once per trial. After dart shooting, snails were nat used again for another 10 days ta al.law the dart ta regenerate (Tompa, 1982). 129 Snails were randomly chosen for trials by either a coin flip or by a lottery system based on their identification number. As the data often did not fit the assumptions for ANOVA, nonparametrie analyses werc performed (Meddis, 1984) unless otherwise noted. When testing specifie hypotheses after completing the initial

. nonspeeif ic 1 analysis, the statistic ~2=tl was m:;ed.

Experiment 1. Determination of Daily Rhythms in

Aeti vi ty, Feeding and Sexual Behaviour

The following experiment examines whether behaviaural responsiveness to food and mates follows a daily rhythm paralleling the daily rhythm in activity.

In many molluscs, including Helix, the level of aetivity ean be quantified by a behé.vioural state score

(BSS) derived from the animal' s posture (Preston & Lee,

1973; Lee & Palovcik, 1976, Tuersley & Mc.Crohan, 1987;

Chapter 5). In Helix, the scale consists of 6 ascenàing levels based on the amount of extrusion from the shell, ranging from eompletely withdrawn, to completely extended and loeomoting (Chapter 5). Total aetivity in the colony at any given time was measured by averaging the B8Ss of all the snails in the colony. BSSs were 130 collected ever'j hour. After inspecting the activity results, the day was divided int"o three periods: the first 3 h of photophase, the rest of photophase, and scotophase; subsequent statistical analyses justified these divisions (see below).

Food consumption was determined by measuring the carrot weights in the 4 group boxes at the beginning and end of each of the 3 time periods, as weIl as at 2 h intervals during photophase. The weights for each time

period were averaged over the 4 boxes for 10 consecutive days, and weights were converted to hourly rates to allow comparisons between each time period. The number of copulations that occurred during each time period in the group boxes was also recorded. During scotophase, copulations were assessed from photographs taken at hourly intervals under red light. As copulation

typically lasts 6 hours (Chapter 2), aIl cases of copulation were recorded. Replicates were found not te differ significantly

(~>O.l, nonparametric factorial analysis, related samples) and therefore the data were pooled for each behaviour.

-- 131 Results

Snails showed a prominent dai1y rhythm in their

activi·... y (Fig. 1). They were most active during scotophase and the first :; hours of photophase Crt2

they were most active (R

measures within each time period (R>O .1, nonparametric factorial analysis, related samples) suggesting that there were no large trends within the time periods. Therefore, there is no justification for any further divisions of the day (Fig. 1).

(~ l' t

1 5 t f .- ~ 0 c 4 t (J) "-...... L en r t- (f) T ~ 3 CD T • [ ! '-" l l l 1 Q) • f > , /1) 2 l T l ,,' • r +J • :~ ~ ~ .... 1 .L l U 1 <{ l photophose scotophase 0 0 3 6 9 12 15 18 21 24 n=26 n=26 n=26 n=26 n=26 n=8 n=8 n=8 n=26 Time (h) Figure 1. The daily rhythm of behavioural state scores CBSSs) in Helix aspersa. Snails were significantly more active during early photophase and scotophase than they were during mid and late photophase (~

'U-

0 c en 4 T '-... ~ • (/) r (/) T CD 3 ~ 1 l •l 1 Cl) > Cl) 2 l r •l .....>. • T • > :.;:; 1 • .L 1 0 1 1 • < photophase scotophase 0 0 3 6 9 12 15 18 21 24 n=26 n=26 n=26 n=26 n=26 n=8 n=8 n=8 n=26 Time (h) ------

Figure 2. The daily rhythm8 of BSSs, food consumption and mating. AlI scores were normalized to the largest data point value for each of the 3 behaviours. A score of 1.00 is equivalent to a BSS of 4.0, a food consumption rate of 0.8 g carrot/h/10 snails, and a copulation rate of 0.15 matings/h/10 snails. There were no significant

differences in the daily rhythm for each acti vi ty (12)0. l, nonparametric factorial analysis, related samples). Matings were assigned to the time period in which copulation began. AlI values are means ± standard error. 1.00 CS] 8SS "0 C [3J Food consumption o CZJ Copulation frequency g' 0.80 "0 (f) Q) Q) Q) 0 lJ... u 0.60 cJi(j') (J) 0'1 CIl C "0 .... Q) 0 0.40 N~ o E o~ 0.20 z

Oh -.3h 3h-16h 1 6 h - 24 h 1///////1/ 1/1 photophase scotophase Time (h) 132 Exper iment 2 . Effects of Serotonin and Handling on Activity Levels

Handling (Kupfermann & Weiss, 1981) and serotonin

(Palovc~k et al., 1982) increase behavioural state in Aplysia. Before the same methods could be used to increase behavioural state in Helix, it was necessary to confirm an equivalent effect in this species.

Snails were randomly assigned to 8 group boxes. Each group was randomly assigned to one of 4 treatments: injections of serotonin (0.1 mL, 5 x 10 -5 M; approx. 5 x 10-7 moles/kg body weight); 0.1 mL of snail saline

(Chase, 1986, the vehicle for serotonin); handlingi or unhandled and uninjected (controls). Handled animaIs were flipped onto their backs, inducing a righting reflex and an increase in activity. Half of the groups received treatments daily at the beginning of photophase,

while the other half received treatments 1.0 h into photophase (late photophase), when snails were typically

inactive (Fig. 1.). 8SSs were recorded immediately before treatment as

well as 5 min and 1 h after treatment.

r ------

133 ... Results

Five minutes after treatment, the average BSSs of the serotonin inj ected snails increased irrespect ive of the photophase in which the injection was given (Fig. 3, R

(Fig. 3, 12<0.01). There was no difference between the saline effect and the handling effect. The effects were largest when the treatments were given during late photophase (the inactive phase; 12<0.05). Five minutes after treatment, snails were equally active (i.e. had similar BSSs) in both early and late photophase. One hour after treatments given in early photophase, handled, saline and serotonin injected snails were not significantly more active than controls. However, 1 h after treatments given in late photophase, serotonin injected snails remained more active than saline and handled snails, which were more active than controls

(~<0.05, nonparametric factorial analysis, related samples) . Nevertheless, snails treated in late photophase were significantly less active 1 h after treatment than were snails treated in early photophase (R

.' 3.5

3.0 o c (/) 2.5

(f1 "'"(f1 CD 2.0 c Cl) 1.5 (/) o Cl) U 1.0 c 0.5

0.0 5HT soline han die control 5HT soline handle control Treatments glven at 0 h Treatments given at 10 h

1 l 134 ( increased activi ty, the magni 'tude and duration of the effect was dependent on the time of treatment.

Experiment 3. Effects of Serotonin and Handling on Food consumption

Serotonin and handling have been shown to increase 5SSs (experiment 2). In the followinq experiments, we

test their effects on the snail t s responsiveness to food stimul i (experiments 3 and 4) and sexual stimuli

( exper iments 5 and 6). If . central arousal' has a direct, positive, effect on both feeding and sexual behaviour, then increases in . central arousal' should result in increases in both feedinq and sexual behaviour.

The snails were randomly reassigned to the 8 qroups described above. The food consumption of each group was determined by measuring the difference in carrot weight from immediately prior to a treatment to 3 h later. The number of snails observed scraping the carrot at least once with their radula was also recorded. The results were averaged over 10 consecutive days. Each snail was

weighed before and after the 10 day trial.

( 135 Results

The effects of serotonin and handling on food intake did not exactly paraI leI their effects on BSSs. Trials given at the start of photophase had rio effE:!ct on food intake, even though activity was significantly increasl3d (Fig. 4, g>O.l, nonpalametric factorial analysis, related measures) . Trials given at late photophase did significantly increase food intake over controls (Non-parametric factorial analysis, related samples, ~

(4.8 ± 1.2), saline (5.1 ± 1.4) and handled (4.5 ± 1.1) groups, but only during late photophase (g

Despite the increase in food intake in late photophase, there was no net weight gain over the 10 day trial in the handled, serotonin or saline injected snails compared to controls (Kruskall-Wallis, R>O.I). Figure 4. The effects of serotonin, saline and handling on carrot consumption. Serotonin, saline and handling significantly increased carrot consumption compared to controls, but only when treatments were given in late photophase (n<0.02, nonparametric factorial analysis, related samples). Scores represent the amount of carrot consumed, per group, averaged over 10 trials. Values are means ± standard error...... 0 'C..... 0.5 ,.,.,..c '-.... (J) 0.4 0 c (J) 0 0.3 '-....- ...... 0\ ----0 0.2 V E :J (J) C 0 0.1 ..,(J 0 ~ ~ 0 0.0 () 5HT saline hondle control 5HT saline han die control Treatments given at 0 h Treatments given at 10 h 136 Experiment 4. Effects of Serotonin and Handling on Bite Frequency

Increases in BSS may have an effect on the speed with which food is eaten even in the absence qf an effect on the amount consumed. Feeding speed is one parameter used to estimate food arousal (Susswein, Weiss & Kupfermann, 1978) • The following experiment examines the effect of increasing the BSS on bite frequency. colony snails were gently placed 1 cm away from a piece of carrot attached to a force transducer. The number of bites (scraping the carrot with the radula), recorded visually, over 10 randomly selected 30 s segments was identical to the number of bites recorded for the same time period on a chart recorder by means of the force transducer. This indicated that the force transducer gave an accu rate measure of bite frequency. prior to each trial, snails were randomly assigned to one of 4 groups, 5 snails per group: injections of serotonin (0.1 mL, 5 x 10-5 M); 0.1 mL snail saline; handled; and unhandled, uninjected (controls). Baseline bite frequencies were determined during the first 5 min. Snails that did not respond to the food stimulus within 30 s were discarded. Snails were then subjected to one of the 4 pre-assigned treatments. Bite frequency was r recorded for the next 5 min, and for 1 min periods every ,

- 137 5 min for the next 15 min. Because injections and handling interrupted feedinq, the 5 min post-treatment trial was not started until the snails had spontaneously resumed feedinq.

Results

Snails averaqed 4.6 ± 0.5 bites per 10 s durinq baseline trials, 4.6 ± 0.6 bites per 10 s after handlinq, 4.6 ± 0.6 bites per 10 s after saline injections, and 4.6 ± 0.9 bites per 10 s after serotonin injections. There were no siqnificant trends in the number of bites as time proqressed after treatment (2)0.1, nonparametric factorial analysis, related samples). Therefore, neither serotonin injections nor handlinq had any siqnificant effect on bite frequency (nonparametric factorial analysis, related samples R>O.l).

Experiment 5. Effects of Serotonin and Handlinq on Sexual Proclivity

In these experiments we examine the effects of increasinq behavioural state on the responsiveness to sexual stimuli. If . central arousal' potentiates responsiveness ta sexual stimuli, then increasinq ( 138 'central arousal' should correlate with increased responsiveness to sexual stimuli. Feeding behaviour is also monitored to examine the interactions between 'central arousal', feeding, and mating. We differentiate

between 2 types of respensiveness: proclivity (the likelihood of a response), and arousal (the intensity of

the response). In this experiment we consider proclivity, in the next, arousal. There is evidence

that, in Helix, sexual proclivity and sexual arousal are

controlled by 2 distinct mechanisms (Chapter 4) . As sexual proclivity increases, the number of snails that display courtship behaviour and which eventually

copulate also increases (Chapter 4). Therefore,

proclivity was estimated from either the number of

courting snails or by the number of copulating snails.

The effect of increasing BSSs on sexual proclivity was

examined by following the same procedure used to test for an effect of increased BSSs on feeding behaviour

(experiments 3 and 4). Treatments were given daily fer 5

days to the 8 groups described above (experiment 3), 10

snails per group. Snails were sexually isolated fer 10

days between 3 replicate trials. The number of copulating snails in each group was recorded for 5 days

befere trials as weIl as during the trial. If increasing

BSS correlates with increased mating, then the (

1, j 139 ....,.. serotonin, saI ine, and handled groups should copulate more than the controls.

In the above test, snails had constant access to mates, except between trials. Because copulation decreases sexual proclivity, their sexual proclivity was progressively declining (Chapter 4). To determine the effects of the 4 treatments on snails with high sexual proclivity, treatments were given to snails that had been aestivated for 2 months prior to trials. Aestivation in H. aspersa increases the likelihood of subsequent mating (Bonnefoy-Claudet & Deray, 1984). Chapter 4 shows that sexual isolation also increases sexuai proclivity. The snails were therefore sexually isolated after being removed from aestivation. At the start of each trial, snails were randomly divided into 8 groups of 12 snails each. During the trials, snails were gently placed into

a Lucite observation box (20 cm x 20 cm x 8 cm) at either

early photophase or late photophase. Thirty minutes after being placed in the observation box, each group was given one of the 4 preassigned treatments: injections of

serotonin (0.1 mL, 5 x 10-5 M); injections of 0.1 mL

saline; handling; or no injection and no handling (controls) . The number of snails that displayed

courtship behaviour was recorded. The snails were

prevented from copulating by returning them to their individual containers when their genital eversions ( 140 reached stage 5. The time required for each snai l to reach dart shooting threshold (stage 5) was also recorded. Because placing the snails in the observation bo:x involved handlillg them, and as this May have influenced the resul ts, a third test was performed in which snails were not handled prior to trials. There were 3 groups: handled, unhandled, and control. Each group contained 12 snails. During the trials, the barriers between snails in the handled and unhandled groups were removed without disturbing them. The control group remained isolated throughout the trial. By comparing the BSSs of the control group and the unhandled group, the effect of conspecific contact on BSSs cou Id be measured. After the barriers were removed, the snails in the handled group were flipped onto their backs. The BSSs of the snails were recorded before treatment, 5 min after, 1 h after and 2 h after treatment. In the handled and unhandled groups the number of snails that courted as weIl as the time required for each snail to reach dart shooting threshold was recorded. The number of snaiis that were observed biting the carrot in aIl 3 groups was aiso recorded. Snails were removed once their genital eversions reached stage 5. The above procedure was repeated during late photophase to determine whether the ( handling effect was dependent on the time of day. - 141 Results

Despite clear effects on activity and feeding (Fig. 2), neither serotonin nor handling had any significant effect on the copulation frequency of snails that were not sexually isolated, irrespect ive of the time of treatment (Fig. 5, R>O.l nonparametric factorial analysis, related measures). Even though snails treated in late photophase were active for more hours of the day than were controls (Fig. 3), the y did not use this time to mate (though they did use it to eat, Fig. 4). Mating tends to inhibi t feeding, and appears to be the more

dominant behaviour (Chapter 5). Therefore, it is unlikely that feeding during late photophase inhibited any putative stimulation of mating behaviour. In sununary, there is no evidence that increasing BSSs can

induce sexual behaviour. However, in sexually deprived snails, mating occurred in both early and late photophase in aIl animaIs (Table 1). There was no significant difference in the number of snails that courted as a function of treatment, during either early or late photophase (Table l, R>O.l, G-test). Nor did the treatments affect the time that snails required to reach dart shooting threshold, again -... regardless of the time of treatment (Table l, R>O.l, Figure 5. The effects of serotonin, saline and handling on copulation frequency. None of the treatments had any effect on copulation frequency (12)0.1, G-test). Each score is the mean copulation rate, per day, averaged over the 10 day trial. Each score represents the median value of the three replicates while the error bars represent the remaining 2 values.

l - - >0- 0 "1j 1.0 ...... CS] number of copulations 5 days prlor to triai .-(/) 0 CZl number of copulations during 5 day triol c 0.8 (/) 0 ..- '-.... (/) c 0 ...... 0 ::J 0- 0 U 1+- 0 \... Q) ..0 E ::J Z 5HT soline handle control 5HT saline handle control Treatments glven at 0 h Treatments glven at 10 h

- Table 1. Serotonln, saline, and handling have no effect on the number of

courtlng snail. or on the tlme required to achleve dart shootlng thresh01d

(ds) ln sexually deprlved snails Each group contained 12 snails. There

vere no significant differences generated by Any experimental treatment

(R>0.1, G·test; R>O .1, nonparametric factori.l analysis, unrelated

samples) .

Treatment

Control Handling Saline Serotonin

Ear1y photophase

time 31.0 ± 24.2 35 2 ± 24.8 42.8 ± 25.9 36.9 ± 23.4

to ds

(min ± sd)

Number of 7 8 7 6

courters

Late photophÀse

Time 35.8 ± 25.6 37.9 ± 22.8 33.4 ± 26.6 38.7 ± 23.5

to ds

(min ± sd)

Number of 8 6 6 7

courters

( ,

- 142 - Kruskall-Wallis). Therefore, increasing BSSs with serotonin, saline or handling did not increase sexual proclivity in sexually deprived snails. When the barriers between sexually isolated snails were removed, thus allowing exposure to conspecifics, the BSSs increased relative to those of continuously isolated controls. However, this increase only occurred when the barriers were removed in late photophase (Table 2, nonparametric factorial analysis, related measures, Q0.1, G-test, R>0.1, nonparametric factorial analysis, related samples). However, handling immediately prior to trials in late photophase had a highly significant effect on the number of snails that courted and on the number that reached dart shooting , Table 2. The effects of removing barriers beeween sexually isolated snal1s, \. and th. depend.nce of thue effec ts on the time of day. Each group

contained 36 sn.ils. Behavioural state scores (BSSs) were taken 1 h after

treatment.

Percentage Percentage BSSjsnail

courting feeding ± sd

Barriers retained

early nia 3.3 ± O.Sa

photophase

late nia 1.8 ± 0.3b phocophasf'

Barriers removed (unhandled) early 72 a 56a 3.5 ± 0.2a

photophase late 23 b 2Sb 2.5 ± o 3c

photophase

Barriers removed (handled)

early 3.4 ± 0.6a

photophase

late 3.3 ± 0 3a

photophase

nia not applicable

Column scores that differ significantly from one another are denoted by

dlfferent letters (percentage courting, 2<0.01, C-test; percentage feeding,

R<0.05, G-test; BSSs, 2<0.05, nonparametric test, related samples.)

( 143 threshold, compared to unhandled controls (Table 2, RO.l, nonparametric factorial analysis), suggesting that these treatments had no effect on sexual arousal. This inference is tested more directly in the following experiment.

Experiment 6. The Effects of Serotonin and Handlinq on Sexual Arousal

Sexual arousal in Helix is characterized by a reduction in courtship duration, an increase in the stage of a snail's qenital eversion and an increase in the time a snail will spend in contact with an anesthetized 144 conspecific (Chapter 4). The effects of serotonin and handling on each of these parameters were examined. In the first test, mating snail pairs were randomly assigned to one of 4 treatments: serotonin; saline; handling; and unhandled, uninjected controls. The genital eversion stage at the time of injection was recorded. The time required for each snail to reach dart shooting threshold was aiso recorded. In the second test, 4 previously isolated snails were placed in an observation box. At preassigned genital eversion stages, aIl snails except the target snail were removed from the box. Target snaiis were randomly given one of 4 treatments as above. The genital eversions at 30 s, 1 min, 2 min, 3 mir and 5 min were recorded. In the third test, snails, at preassigned stages of genitai e·/ersion, were confronted with an anesthetized

partner placed at a distance of 1 cm. The procedure has been previously described (Chapter 4). The time that snails remained in contact with the conspecific du ring the 5 min trial was recorded. Results Increasing snail activity by injecting serotonin or

saline, or Ly handling the snail~, did not reduce courtship duration (Fig. 6, g>O.l, Kruskall-Wallis). None of the treatments significantly increased the stage ( of the snail's genital eversion (g>O.l, G-test), although •

Figure 6. The effects of serotonin, saline, and handling on the latency to dart shooting, and their dependence on the stage of genital eversion at the time of treatment. There was no effect of any of the treatments at any geni tal eversion stage (R>O. 1, Kruska11-Wall is) . Each score represents the average latency to dart shooting for 10 snails. Values are Median ± 1 quartile. Snails that failed to mate after courting received a maximum score of 60 min.

- (

60 serotonin injected CJl ~ c CXJ saline injected ~ 0 ~ handled 0 oC CS] control (f) 40 ~ ,...... ~o .-c u -..-E 0 ~ >. u 20 c ....,Il) 0 ....J

0 0 1 - 2 3 - 4 Genital everSlon stage at injection Figure 7. The effects of serotonin, saline, and handling on the time spent in contact with an anesthetized conspecific, and their dependence on the stage of genital eversion at the time of treatment. Treatments had no effect on contact time at any stage of genital eversion (R>O .1, Kruskall-Wallis). Each score represents the average of 10 snails. Values are median ± 1 quartile.

) ( ~c 4.0 E 3.5 ~ serotonin injecterl .c ---u [X] saline injeded ._4-J .-4- ~ 'ü 3.0 ~ handled Q) 4-J U 0.. CSJ control 0 en 2.5 4-J c C 0 o u u 2.0 \J C Q) .- N 1.5 (l):;:; E 0 ~ 1.0 .-1- ...... Cf) Q) c 0.5 CJ 0.0 0 1-2 3 - 4 n= 1 0 n=9 n= 11

Genita 1 eversion stage at injection 145 of the snail's genital eversion (2)0.1, G-test), although saline and serotonin sometimes increased the size of the

penial lobe. This change may have been due to increased

liquid volume in the hemocoel, and hence increased blood

pressure in the cephalopedal area. None of the treatments increased the time snails spent in contact with an anesthetized conspecific (Fig. 7, 2>0.1, Kruskall-Wallis) . Therefore, there is no evidence that either serotonin or handling increased sexual arousal.

Discussion

Helix exhibits a variability in its responsiveness to

food and mates which follows a daily rhythm paralleling

that of activity (Fig. 2). This finding superficially

supports the hypothesis that a 'central arousal' system positively modulates activity, feeding and sexual

behaviour. However, in Helix, particularly in the case of sexual behaviour, the effect of 'central arousal' appears to be only a permissive one, allowing the expression of sexual proclivity and arousal without directly influencing the mechanisms underlying sexual proclivity and sexual arousal. This conclusion for Helix contrasts with the case in Aplysia, where the mechanism(s) underlying 'central arousal' is thought to .--_... ------

Figure 8. Schematic summary of the relationships between

'central arousal', feeding and sexual behaviour. BSS was used to estimate 'central arousal'. The boxes labelled 'sexual procl i vi ty' and 'feeding procl i vi ty' represent the sum of the neural and hormonal factors that contribute to creating the change in respons i veness to sexual or food stimuli. The boxes labelled 'mating' and , feeding' represent the neural and hormonal components

responsible for the execution, and the speed of execution (Le. arousal), of either feeding or sexual behaviour. Al though the boxes are separate , sorne of the same components may contribute to both proclivity and arousal

for either behaviour. 'Central arousal' is shown as having a permissive, or gating, effect on sexual behaviour, but it has no direct effect on sexual

proclivity. Sexual proclivity leads to se:{ual behaviour only when sexual stimul i are present and when the snail is at some minimum level of 'central arousal'. Whether increased 'central arousal' can directly affect the components responsible for determining feeding is not known. Arrowheads represent exci tatory relationships; a

filled circle represents a relationship that is

predominantly inhibitory. Question marks denote plausible, but not yet demonstrated, relationships. r 1.- ... Feeding r- Mating r+-

? Food ~? ~'l' Sexual Stimuli Stimuli ? " Sexual 4 Feeding Proclivity Procllvlty i!"

--...... Central Arouea. r"

'1 146 positively modulate the mechanism(s) responsible for proclivity and arousal in both feeding and sexual

behaviour (Ziv, Benni, Markovich, & Susswein, 1989). Although behavioural state was used to estimate the

relative level of 'central arousal', the 2 terms are not synonymous. 'central arousal' is characterized by behavioural responsiveness as weIl as by activity. Behavioural state can be used as an estimate of 'central arousal' only if it is assumed that behavioural state never increases except when • central arousal' also increases. Given that increases in behavioural state do seem to correspond to increases in behavioural

responsiveness (Preston & Lee, 1973; Andrew, 1974; Chapter 5) the assumption appears to be val id.

• Central arousal', as estimated by the BSS, plays a permissive role in that it is a necessary precondition

for mating. In terrestrial snails, like Helix, high

cephalopedal blood pressure is presumably required not only for locomotion and feeding, but also for such

courtship behaviours as dart shooting and penile extrusion (Dale, 1973). High cephalopedal blood pressure

results in extrusion from the shell (Dale, 1973). since extrusion from the shell is equivalent, by definition, to

a high BSS, a high BSS is required before mating can occur. However, activity is not a sufficient precondition for mating because increasing activity has 147 no effect on copulation frequency (Fig. 5). The inability of an increase in behavioural state to increase copulation frequency is contrary to what would be

expected if a 'central arousal' system directly modulated sexual proclivity. However, it is consistent with the idea that 'central arousal' is simply cerrelated with the necessary physiological conditions for rnating. In apparent contradiction to the above, increasing activity in late photophase led to increases in both feeding and mating behaviour in sexually deprived snails (Table 2). However, copulation rates in these snails were identical to those found in the early photophase group (Table 2). Assurning snails with sexual proclivity were equally likely to be in either the early or late photophase group, then increasing 'central arousal' did net increase proclivity above that already present and capable of being expressed when the snails were naturally

active in early photophase. Therefore, it appears that increased levels of 'central arousal' did not result in increased rnating behaviour, but merely allowed for the expression of t.he underlying sexual proclivity. It is likely that sexual behaviour requires specifie internaI conditions before it can be initiated (Chapter

4) • Moreover, evidence suggests that changes in these

,,; conditions occur only over a period of days (Chapter 4), 1 148 not minutes, as might be expected if increasing activity could induce sexual behaviour.

Although the rnechanism(s) responsible fo~ 'central arousal' appear ta have no direct effect on sexual proclivity, snails isolated sa as to increase their sexual proclivity, increased their BSS when exposed to sexual stimuli in late photophase, a time when they are usually inactive (Table 2). Exposure ta food causes a similar increase in BSS in hungry Helix (Chapter 5). Food stimuli can also increase BSS in food deprived Pleurobranchaea (Palovcik et al., 1982). Therefore, a positive proclivity for either food or sex can, in the presence of the appropriate stimuli, induce an increase in 'central arousal'. Sexually deprived, unhandled snails are less active when exposed to conspecifics in late photophase than when

they are exposed in early photophase. Moreover, sexually deprived snails mate during late photophase less

frequently than they do in early photophase (Table 2). Therefore, only a proportion of sexually deprived snails are capable of mating during late photophase. During early photophase, when 'central arousal' is high, lower levels of sexual proclivity are required for sexual stimuli to elicit courtship. This may explain why more

snails mated in early photophase and why the rhythm of - 1

149 mating in Hel ix followed the dai1y acti vi ty rhythm (F ig .

2) • Increasing activity not on1y fai1ed to affect sexual proclivity, but it also had no effect on sexual arousal (Fig. 6, Fig. 7, Table 1). • Central arousal' also has no effect on bite frequency, which was the only parameter of food arousal that was measured. A lack of effect on sexual arousal and feeding arousal would be expected if a 'central arousal' system has a permissive effect, and not an active effect, on the neural mechanisms that control these behaviours. The evidence from the experiments on feeding behaviour is consistent with the hypothesis that 'central arousal' serves a permissive function for feeding, similar to that for sexual behaviour, though the ev idence does not exclude the possibility that 'central arousal' exerts a more direct effect. We have already reported that latency ta bite is negatively correlated with BSSs (Chapter 5). In the present study, we found that

increasi~g activity increased feeding (Fig. 5), and snails in late photophase were bath less active, and fed 1ess, than snails in early photaphase (Figs. 3 and 4). These results support the hypothesis that a 'central arousal' system directly modulates the system responsible for feeding proclivity. During late photophase, however, snails are inactive and incapable of eating. As with 150

1 ... sexual behaviour, it is possible that increasing activity during late photophase simply allowed the expression of feeding proclivity, rather than directly affecting feeding proclivi ty. BSSs and feeding will appear as tightly coupled behaviours if feeding proclivity is normally present and satiation is rare. It is paradoxical that increasing activi ty had an effect on food consumption only in late photophase (Fig. 4). Possibly, the initial increase in activity after treatment was due to a . stress' or . startle' response. There is evidence that tactile contact initially inhibits

feeding in Helix (Everett, Ostfeld & Davis, 1982), and serotonin, saline and handling treatments aIl involve tactile contact with the animal. Once this effect decays, snails are no more active during early photophase trials than are controls (experiment 2). However, during late photophase, snails are . aroused' from an inactive state. Snails may initially be 'stressed' or 'startIed'

as during early photophase, but as this effect decays,

the snails slowly return to dormancy. As the snails return to dormancy, they pass through levels of activity that are compatible with eating. Therefore, when compared to completely inactive controls, they exhibit an increase in food consumption • A similar . stress' or

. startle 1 phenomenon is thought to occur in Aplysia,

which shows decreased feeding immediately after handl ing 151

but increased feeding 4 min later (Kupfermann & Weiss,

1981) •

Fig. 8 summarizes the interactions between 'central arousal' (estimated by BSS) , feeding and sexual behavior. It should be noted that a 'central arousal' system has yet to be identified in any mollusc. Nevertheless, some mechanism(s) must exist to account for the coordination of behavioural responses to diverse internaI and external

stimuli (Kupfermann & Weiss, 1981). Whether this system exists as a set of neural/neurosecretory cells with a special i zed 'central arousal' function, or whether 'central arausal' exists by virtue of the parallel connections of sensory cells and endogenous oscillators

to each of the different behavioural centres, remains ta

be discovered. A recent report, however, de scribes neurons which appear to be specialized for exerting a

. motivational effect' when activated (Teyke, Weiss, &

Kupfennann, 1990).

In contrast to Helix, in Aplysia bath sexual behaviour and feeding appear to be positively modulated by a 'central arousal' system (Ziv et al., 1989a). Food

deprivation decreases the amount of tirne Aplysia spends immobile and increases mating activity and spontaneously occurring appetitive feeding behaviours (Susswein, 1984;

Weiss, Koch, Koester, Rosen, & Kupfermann, 1982). In Helix, however, food deprivation decreases both BSS and _H 152 mating activity (Chapter 5). The differences in the interactions between activity, sexual behaviour and feeding in these 2 gastropod molluscs are interesting from an ethological viewpoint and underscore the importance of comparative studies. Aplysia and Helix have very different reproducti.ve strategies. Aplysia lives only one reproductive season, and during the breeding season spends over 25% of its time mating (Ka,ldel, 1979; Susswein, et al., 1983). It produces ten::. of thousands of eggs in a single bout of egglélying (Kandel, 1979). Helix, on the other hand, typicaJly lives more tha', one reproductive season, but even with surplus food and access to mates, these snails mate less than once every 28 days during the breeding season (potts, 1975; Moulin, 1980). Helix lays less than a hundred eggs per oviposit (Potts, 1975). Therefore, for Helix, which feeds far more often than it mates, a system that enhanced responsiveness to sexual stimuli whenever feeding behaviour was facilitated, would be inappropriate. The feeding and reproductive strategies of Helix require that feeding and sexual behaviour be capable of independent modulation. The different reproductive strategies of Helix and Aplysia appear to have led to differences in how sexual and feeding behaviours are related to 'centr.al arousal'. 153

other molluscs also differ from ~plysia in the way their behaviours are related to 'central arousal'. In Lymnaea, handling has no effect on food arousal (Tuersley and McCrohan, 1987). Again, this difference is thought to be a reflection of the differences between the feeding strategies of the two species (Tuersley & McCrohan, 1987: Tuersley, 1989). Lymnaea is a herbivorous grazer which feeds continuously, and satiates rarely. Aplysia feeds in bouts, and altering the level of food arousal is one mechanism by which the duration and intensity of the bouts can be controlled. Such control is not required in Lymnaea.

Although serotonin had a significant effect on activity, it was not more effective than injections of saline or handling in potentiating feeding and mating. Nevertheless, Kemenes and S.-Rozsa (1987) have evidence

that serotonin is involved in food arousal in Helix. It

is possible that injections, being invasive, and handling, which may be , stressful ' , activate the 'central arousal system' postulated by palovcik et al. (1982). This may make the injected serotonin

superfluous. It is also possible that systemically deI i vered serotonin does not mimic the effect of serotonin in vivo in the sexual and feeding neural

systems. The role of serotonin in 'central arousal', - 154 feedinq and sexual behavieur in Helix remains te be determined.

( ------....._------

155

Chapter 7. Thesis Discussion 156 sUmmary and model of Helix aspersa sexual behaviour

Fig. 1 illustrates a schematic model of Helix sexual behaviour that is consistent with the results presented in the previous chapters. The following sections give a summary of Helix aspersa sexual behaviour based on this model.

Sexual proclivity In a constant environment with moderate tempe ratures , abundant food and a summer light cycle, sexual isolation, aestivation, or simply the passage of time since the last copulation, will increase a snail' s sexual proclivity

(Chapters 4 and (5). The increase in sexual proclivity is . " graduaI and occurs over several days, suggesting that

i there is a hormonal component in its control. 1 '1 '1 1. Sexual proclivity cannot be induced by increasing ," .

1 behavioural state (BS) (Chapter 6). This suggests that a ~ central 'arousal ' system responsible for increasing BS !/ '1 (modelled in Fig. 1) has only a permissive effect on j 1 sexual behaviour. In other words, it does not interact directly wi th the mechanisms controlling ei ther sexual arousal or sexual proclivity (Chapter 6). However, in the presence of sexual stimuli, snails that have been

sexually isolated for several days will increase their as, if it is below the level needed for locomotion r/ ,~ t r ....

1> 1., ~ ~ t \

Figure 1. Schematic model of Helix aspersa sexual

behaviour emphasizing its relationship with 'central arousal'. Shaded arrows denote excitatory relationships while shaded circles denote relationships that are inhibitory. Open arrows signify that both inhibitory and

excitatory relationships are possible. Points at which two or more inputs converge denote that aIl inputs are

required before there is an effective output .

. ,. .'-,. (

Locomotion Internai and Fe.dlng External Stimuli etc.

BS Sexuel SpeCifie Central Arouaal Central Arousel

Sexuel Stimuli

Aestlvatlon

Sexual laolatlon Sexual Procltvlty Sexuel Aroua.'

Copulation Sexuel Stlmul Dlgltlform Gland Mucus

{ 157 (Chapter 6). Under these conditions, sexual stimuli are capable of activating the BS central arousal system (BSCAS) . Gomot and Deray (1987) list various physical factors such as humidity, temperature, food availability, time since previous aestivation, time since previous copulation, and photoperiod, that appear to be important in Helix in controlling sexual proclivity. Of these factors, the time since the previous aestivation appears to be the most important, followed by the time since last copulation (personal observations). Humidity, temperature, and food availability are likely to be permissive factors, but are unlikely to interact directly with the mechanisms involved in sexual proclivity. Snails must be active to court, and physical conditions that promote aestivation obviously prevent courtship. The effect of photoperiod on sexual proclivity is not completely clear. Despite the evidence that a long photoperiod cues sexual proclivity in this species (Bonnefoy-Claudet et al., 1987), in California, Helix aspersa tends to mate and reproduce during short, not long, light periods (Ingram, 1947; Potts, 1975).

Moreover, Guemene and Daguzan (1982) ha'/e found that Hel ix will breed in both constant 1 ight and constant darkness, indicating that photoperiod is not a necessary cue for the development of sexual proclivity. 158 1 , Sexual proclivity appears to have a circannual cycle (Bailey, 1981), therefore given favourable conditions, its ons et may be controlled by an endogenous clock.

Sexual proclivity declines after copulation. This decrease is not dependent on spermatophore exchange (i.e. on factors from the dissolving spermatophore) (Jeppesen,

1976) . Dart shooting alone will decrease sexual proclivity (Jeppesen, 1976; Lind, 1976). The actual performance of dart shooting is not required for the

decline of sexual proclivity; the same decline in proclivity is seen in snails that are missing dart sacs

(Jeppesen, 1976). Snails that are missing dart sacs perform 'phantom' dart shooting, suggesting that the

neural circuits involved in dart shooting continue to function despi te the loss of the dart sac (Chapter 3). This suggests that the decline in sexual proclivity may

be triggered directly or indirectly by the neurons in the

mesocerebrum which activate the firing of the dart.

The frequency of sexual proclivity differs widely

amongst molluscs. In Helix, sexual proclivity occurs

rarely. It has a time course very different from that of feeding which occurs more or less daily (personal observations) . The estimates for mating frequency in

Helix vary between different reports. Guemene and

Daguzan (1982) found that snails mated an average of once

every 15 days. They checked the snails for mating 159 behaviour twice a day for 8 weeks during the breeding

season. Moulin (1980) found that snails rnated an

average 01: once every 30 days over the entire 5 month breeding season. In this study, snails were checked once

per day. Madec and Daguzan (1987) found that their

snails mated once every 30 days, and snails were checked

twice per day for 15 weeks during the breeding season.

Chapter El reports that the snails used in this study

mated an average of once every 12 days, and the snails

were mon ltored hourly over the 24 h cycle for 3 weeks after being rernoved from aestivation. Despite the variation amongst the estimates, aIl of these studies

show that Helix courtship and copulation is a rnuch rarer event thdn it is in the marine mollusc, Aplysia (Susswein

et al., 1983). This naturally raises the question of why periods of sexual proclivity should occur more often in Aplysia than

in Heli:~. Though the ecological information on both species is rather scanty, one can speculate on sorne of

the forces that may have contributed to the difference.

In ~sia, frequent mating has been hypothesized to

increasE~ the genetic variabili ty of offspring. Given their highly changeable habitat, this increase is thought to be aclaptive (Susswein et al., 1983). However, Helix tends to sp~nd its entire life in small colonies with virtually no immigration (Potts, 1975). Selander and 160 Kaufman (1975) estimate that the deme size for Helix is about 15 individuals. Therefore, highly promiscuous mating in Helix may not substantially increase the variability of its offspring, even if variable offspring

were also advantageous for Helix. Perhaps it is not surprising then that snails do not seem to have evolved

any behavioural mechanisms to encourage outbreeding.

Snails appear to mate at random, and neither favour nor

disfavour newly introduced individuals from a different populatjon (Woyciechowski and Lomnicki, 1977). Neverthe1ess, as Baur (1988) points out, fertilizing

other snails decreases the chance that a snail will leave no genetic offspring if i t is preyed upon. However,

because the cost of mating is not known, it is difficult

to deterrnine what the optimal rate of copulation may be

for Helix. The co st of copulation may be especially high in Helix

aspersa because of the concomitant risk of parasitisme

In Helix aspersa there are a number of parasites that are

spread exclusively by sexual behaviour (Lind, 1973;

Morand, 1988). At least one of these parasites, the

nematode, Nemhelix bakeri decreases egg output (Morand,

1988) . Although the rate of parasitism is not known, this pressure may have contributed to the low mating

rate of Helix, as low mat ers would have a decreased { chance of being parasitized. As Helix can store viable 161 sperm for 4 years (Taylor, 1900) frequent copulation may not be required for a typical season's egg production. If the physiological basis of sexual proclivity could be elucidated in both species, it would be interesting to compare how the control of sexual proclivity has been shaped by evolution to give the 2 species their different rates of copulation.

Sexual arousal

In snails with sexual proelivity and a relatively high BS, sexual stimuli lead to courtship behaviour (Chapter 6). Sexual arousal does not oceur unless the snail can engage in tactile contact with a conspecific. An anesthetized conspecific can induce sexual arousal but it is less effective at inducing arousal in its partner than is an active snail (Chapter 5). A rubber snail model is ineffective in inducing sexual arousal. Attempting to induce sexual arousal by stroking the animal with a brush is equally ineffective (Syzmanski, 1913- discussed by

tind, 1976). This suggests that a combination of the appropriate tactile and chemosensory stimuli are required to elicit sexual behaviour in Helix. Visual cues are unlikely to be important because Helix has poor forro vision (Hamilton and Winter, 1984).

,.. , .. 162

Sexual arousal is probably partly controll~d by a positive feedback mechanism, initiated by 'tactile and chemosensory eues. As sexual arousal increases, the size of the geni tal eversion increases, which increases the amount of tissue available for stimulation (Chapter 4). This would increase the amount of incoming sexual sensory stimuli. The dependence of sexual arousal on this type of positive feedback mechanism would help to explain why courtship in Helix relies so extensively on mutual lip genital contact (Chapter 2). This reliance on sensory stimulation illustrates another fundamental difference between sexual arousal and sexual proclivity. Sexual arousal is controlled predominantly by exogenous stimuli, while sexual proclivity is more dependent on internaI eues. This again suggests that each is controlled by a separate mechanism. As weIl as tactile and chemosensory stimulation, snails also use a courtship pheromone found in the

,1 digitiform gland mucus to increase the sexual arousal of a partner (Chapter 3). During dart shooting, the dart carries the digitiform gland mucus into the body cavity of the recipient. Presumably the pheromone is then

dispersed by the circulatory system to i~s target tissues. 1

"'"' 163 - Sexual arousal exerts specific effects on other behaviours such as feeding and locomotion (Chapter 5). Since these effects do not occur with sexual proclivity, presumably activating the neural circuits responsible for sexual behaviour coincidentally influences circuits controlling feeding and locomotion. These effects are modelled as occurring via the sexual specifie cem:ral

arousal system (Fig. 1). A similar phenomenon is observed in Pleurobranchaea. In Pleurobranchaea chemostimulation alone does not inhibit withdrawal behaviour; the animal must actively feed ( i. e. exhibi t food arousal) to inhibit withdrawal (Davis, et al., 1977). The level of arousal determines the strength of i ts effeets on other behaviours (Chapter 5). This, too, is observed in Pleurobranchaea (Kovac and Davis, 1980b). This implies that the sexual specific central arousal system must be capable of a graded effect.

Proclivity and arousal in molluscs

~~g. 1 shows sexual 'arousal' divided into 2 components, sexual proclivity and sexual arousal. Chapters 4, 5, and 6 illustrated the utility of this separation. Given that the 2 components are likely to be mediated by separate mechanisms, the se3rch for the 164

physiological basis of 1 arousal' would probably benefi t by the division being made explicit in motivational studies done in other molluscs. The following section summarizes work on both sex and feeding in molluscs, but restates the results in terms of proclivity and arousal to demonstrate the general applicability of the 2 terms.

Sexual 'arousal'

The functional division of sexual 'arousal' into separate components is weIl established in vertebrates and these distinctions appear to reflect real mechanistic divisions in both females (Beach, 1976; Aous et al., 1988) and males (Oomura et al., 1988). These differences have been included in models of rat male sexual behaviour as different types of 'arousability' (Toates and o 'Rourke, 1978). Proclivity

is modelled as the f state of arousability' and is controlled by the level of sexual satiation (Toates, 1986) • In Aplysia, too, it is likely that sexual proclivity and sexual arousal are to sorne extent separate entities. Aplysia has a courtship pheromone that appears to effect sexual proclivity, but there is no evidence that it affects sexual arousal (Audesirk, 1977; painter et al., (. 1989) • 165 As in Helix, sexual proclivity can be estimated in Aplysia dactylomeda by the number of previous non-sexual contacts between animaIs (Zaferes et al. 1 1988). This suggests that, as in Helix, sexual proclivi ty exists prior to any sexual stimulation.

However 1 sexual proclivity and sexual arousal may be less distinct in sorne molluscs than they are in Helix.

In LymX@J~.,g" for example, higher levels of sexual proclivity also correlate with higher levels of sexual arousal (Van Duivenboden, 1984).

Food 'arousal '

The division between feeding proclivity and food arousal has been made implicitly in Aplysia (Kuslansky et al., 1987), in which feeding proclivity has been modelled as hav ing a control mechanism that is separate from that controlling food arousal (Weiss et al., 1982). The control of feeding proclivi ty is not addressed in the papers which examine food arousal (i. e. Weiss et al., 1982), suggesting that many molluscan workers see food arousal as separate enough from feeding proclivity that the 2 can be studied in isolation. Nevertheless, the fact that the 2 aspects of feeding are distinct behavioural phenomena, likely ta be subserved by different physiological systems, is never made explici t. 166 In the molluscan feedinq systems studied, however, the division between arousal and proclivity is less complete than it is in Helix sexual behaviour. For example, qut stretch in Aplysia not only decreases feeding proclivity, but it can both increase and decrease food arousal depending on the amount of stretch (Susswein et al., 1978; Susswein et al., 1984). In Limax, gut distension also decreases both feeding proclivity and food arousal (Reingold and Gelperin,

1980). In these systems, both arousal and proclivity utilize similar sens ory information. Nevertheless, a separate term for responsiveness to food stimuli, such as proclivity, would still be useful for studies on feeding. Presently the term satiation is somet imes used to differentiate between the responsiveness to food stimuli and the intensity of the response (Susswein et al., 1984). However, the term becomes cumbersome when trying to describe the state of an unsatiated animal. For example, feeding proclivity in Pleurobranchaea is increased by non-food stimuli such as serotonin inj ections (Palovcik et al.., 1982). If the term satiated were used to describe these results, we would need to say that the animal became more unsatiated after the serotonin injection.

Moreover 1 satiation in the sense of qut stretch does not control feeding proclivity in aIl molluscs. Haminoea 167 zelandiae respond to food continuously, and as intake increases over the processing limits of the digestive system, undigested food is excreted out the anus (Kohn, 1983) . Therefore changes in food responsiveness are unlikely to involve factors that would be well described by the term satiation. Lymnaea, too, appears to respond to food almost continuously, though they were only tested for their responsiveness to food during the day (Tuersley and McCrohan, 1987). In both species, feeding proclivity is a better term than satiation to describe any variations in their food responsiveness. In Achatina, release of gut distension does not resu1t in increased feeding proclivity as might be expected if feeding proclivity were controlled by a simple negative feedback system such as is envisioned controlling feeding in Pleurobranchaea (Davis et al., 1977). Two da ys of starvation has little effect on the snail' s responsiveness to food, and responsiveness is not increased until after 7 days of food deprivation. However aIl feces have passed from the animal by the second day of food deprivation, therefore the gut is unlikely still to be extended after 2 days of starvation

(Croll and Chase, 1980). Presumably feeding proclivity in Achatina is controlled by other internaI processes, with gut distension being only part of the control ,'. system. Even in Aplysia, other factors, such as blood 168 glucose concentration, contribute to the control of feeding proclivity (Bentivegna et al., 1988). These results suggest that it is possible that satiation (the decline in feeding proclivity) is controlled by a mechanism separate from that which ini tiates feeding (increases feeding proclivi ty) . By giving separate terms to arousal and proclivity, these various hypotheses can be more clearly stated. Sorne of the neural mechanisms responsible for changes in feeding proclivity have been studied in Aplysia, as weIl as in otner molluscs. In Aplysia, one group of

cerebral mechanoreceptors have been found which excite the motorneurons for feeding. Various endogenous substances, such as serotonin, FMRFamide, SCPb' buccalin, and myodulin have varying effects on spike width (and hence, potentially on neurotransmitter release) of the mechanoreceptors (Rosen et al, 1989a). These substances are likely to be involved in modulating food responsiveness, and hence proclivity. In Tritonia, feeding proclivity is low during an escape swim. Audesirk and Audesirk (1980) found that during an escape swim, complex mechanoreceptors become less responsive to food stimuli. Work on Pleurobranchaea has also found neurons that are responsible for feeding. As in Aplysia, these cell' s responses are modified by ( 169 1· ,. decreases in feeding proclivity (Davis and Gillette, 1978; Gillette and Gillette, 1983).

The physiological control of food arousal is better understood than that underlying feeding proclivity. In Aplysia, and other molluscs, neurons involved in food arousal have been identified (Pentreath et al., 1982).

The best studied system is that of food arousal in

Aplysia. In this animal the control of bite speed and

bite magnitude is intricate, involving both central and

peripheral mechanisms (Weiss et al, 1982). Food arousal is not controlled by any one cell (Rosen et al., 1989b), but is a highly distributed system. However sorne cells

appear to be dedicated to an 'arousal' function. The

cell CPR probably affects the activity of 2000 other

neurons and can also affect heart rate, as weIl as other

aspects of food arousal (Teyke, et al. 1 1990). It

appears to be involved in coordinating the animal's whole

response t~ facilitate feeding behaviour.

There fore 1 in Aplysia, feeding proclivity is controlled primariIy by gut distension and blood glucose

levels (internaI signaIs), while food arousal is

initiated by sensory stimuli and is mediated primarily

by specialized cells like CPR and MCC. This is similar to one of the differences between sexual proclivity and 170 î '... sexua1 arousal in Helix, in that sexual arousal is dependent primarily on external stimuli, while sexual proclivity is cued by internaI signaIs. By dividing 'arousal' into arousal and proclivi ty, possible relationships between different behaviours rnay

appear more clear. For exarnple, Susswein (1984) and ziv et al. (1989b) found that food deprivation, which increases feeding proclivity, positively rnodulates sexual l proclivity. Sexual proclivity, induced by sexual isolation, has no positive effect on feeding, while sexual arousal increases feeding proclivity. The possible rnechanisms of these interactions are less obvious if proclivity and arousai are lumped into a single entity.

'Central arou;,al' in Helix l' In Fig. 1, the concept of 'central arousal' has been divided into 2 separate components. Based on the resul ts from this work and from others in the molluscan

literature, it seems likely that these 2 components are mechanistically distinct. 'Central arousal' in the sense of a variable describing the relative levei of the animal's

1 alertness' , is often estimated by behavioural state (. (BS) (Lee and Palovcik, 1976). In Helix it has a gating 171 or permissive effect on sexual proclivity (Chapter 6). Behavioural state fluctuates in a daily rhythm and can be increased by both handling and serotonin (Chapter 6). It does appear to have a global, if indirect effect on other behaviours such as sex and feeding (Chapter 6) .

It is conceivable that this type of 'central arousal' is controlled by a unitary system, that is a system in which the components are primarily dedicated to that particular function. This system would receive inputs from the circadian pacemaker as well as other internaI and external stimuli (Le. noxious stimuli). Its output would influence heart rate (necessary to generate the blood pressure needed to increase BS), body wall tone,

and perhaps locomotory circuits as well. In Helix, BS is decreased by mucus from conspecifics

(Cameron and Carter, 1979; Dan and Bailey, 1982). This leads to a decrease in their food consumption and

reproduction (Cameron and Carter, 1979; Dan and Bailey,

1982; Lucarz, 1984). The most parsirnonious explanation

for this effect is that the mucus negatively modulates the system th::lt controls BS in Helix.

The second, and probably rnechanistically distinct,

meaning of 'central arousal' is also shown in Fig. 1-

This type of ' central arousal' consists of a system which ..",. mediates the effects of a behaviour, like feeding, on ....,;. 172 other behaviours and autonomie functions to insure the animal 's abi1ity to perform the activity in a coordinated fashion. It may be mediated by specialized cells like

CPR in Aplysia (Teyke, et al., 1990). Each behav iour will need a different orchestration of suppressed and fac i lita ted behavioural responses. These behav iour-

specifie arousal systems (BSAS) are like1y to be unique. This suggests that each behaviour would have its own system (Leonard and Lukowiak, 1986). Indirect support for the existence of different BSASs for each behaviour in Helix cornes from the finding that it has at least 5 endogenous compounds that can affect

heart rate (Lloyd, 1982) If all • arousing' effects were mediated by the same system, only one or two cardioactive substances would be required.

'Central arousal' in the molluscan literature

The subsequent section summarizes the current theories

concerning 'central arousal' in molluscs. It shows that much of the controversy surrounding different theories of

'central arousal' is largely due to the ambiguities in the terminology. This confusion would be lessened by adopting the practice of spli tting 'central arousal' into 2 separate components, as has been done in the model of { the sexual behaviour of Helix aspersa (Fig. 1). 173 Weiss et al. 's (1982) model of 'central arousal' refers to 'central arQusal' only in its behaviour

specifie sense. , Central arousal' is the neural/neurohormonal system which coordinates the animal' s other behaviours and autonomie systems so that feeding can occur. It is shown as activating the executive feeding arousal system, (equivalent to the food arousal system in this thesis' terminology). The place

of a BS central arousal system (BSCAS) is not explicitly modelled, though the feeding specifie central arousal system is shown as integrating a variety of inputs, one

of which could be information dealing with 8S. weiss et al. (1982) do not imply that their feeding specifie central arousal system mediates sexual specifie central

arousal, or the effects of any other behaviour. It is specifie for feeding.

palovcik et al. (1982), on the other hand, clearly

mean 'central arousal' in the sense of behavioural state.

As with virtually aIl work on molluscs, motivational

variables are constructed with the explicit hypothesis that they are mediated by a real phjsiological system.

Palovcik et al. (1982) propose, "Thus, by activating a central neuronal mechanism, serotonin may bring about the changes associated with the more active behavioral states, namely muscle tone, locomotion, and spontaneous ,-, behaviour, as weIl as increased responsiveness to 174

external stimuli such as food stimuli." (p. 391). In palovcik' s BSCAS hypothesis, unlike the BSCAS hypothesis for Helix, the BSCAS system is thought to directly influence the mechanisms controlling a variety of behaviours.

Kupfermann and Weiss (1981) also postulate the existence of a 'central arousal' system in the behavioural state sense, except that they do not hypothesize that it is necessarily involved in BS. Instead they see this system as positively modulating a variety of behaviours, but it may not necessarily be invol ved in increasing BS. The system is activated by diverse sensory stimuli.

Leonard and Lukowiak (1986), however, deny the existence of a BSCAS. They argue that each behaviour, which they define functionally, is controlled by its own 'drive', and mediates its effects via its own separate system. Therefore they only admit the existence of behaviour specifie arousal systems (BSASs). In support of their hypothesis, they have found that although both food satiation and copulation suppress the gill withdrawal reflex (GWR), the suppression is mediated by a different peptide in each case (Colebrook et al., 1984).

Leonard and Lukowiak (1986) are opposed to the hypothesis that different behaviours would share the same

BSAS. Notice that neither Palovcik et al (1982), nor 175 Kupfermann and Weiss (1981) are suggesting this. Kupfermann and Weiss (1981) and Palovcik et al. (1982) are suggesting that there exists a system which can modulate both the proclivity and subsequent arousal of a variety of different behaviours. Leonard and Lukowiak (1986) would not deny the obvious importance of circadian rhythms in behaviour, but their hypothesis does not deal with behavioural proclivity, as do the postu1ates of Kupfermann and Weiss (1981) and Palovcik et al. (1982), but instead with what mediates behavioural effects. To add to the confusion, the hypotheses of Kupfermann and Weiss (1981) and Palovcik et al. (1982) could be redefined in such a way that there would be no disagreement between their hypotheses and that of Leonard and Lukowiak (1986). If Kupfermann and Weiss (1981) define that the positive modulation of different behaviours by noxious tactile stimul i is a stress behaviour, contro11ed by its own 'drive', and if Palovcik et al. (1982) hypothesize that activity (high aS) or inactivity (low BS) is a1so a functional behaviour controlled by a 'drive', then these theories will now agree with the general hypothesis of Leonard and Lukowiak (1986). Their ideas would clash on1y if someone suggested that the BS 'drive' uses the sarne system as the stress ' drive' • The differences between the various theories would be clearer if a distinction between the 176 two forms of 'central arousal' were made, and if the distinction between proclivity and arousal were adopted.

Heart rate and a BS central arousal system in molluscs

Many workers have suggested that the increase in heart rate that is observed as the animal becomes more 'active' is evidence of a general 'arousal' state (Dieringer et al., 1978; Koch and Koester, 1982; Koch et al., 1984). However, because molluscs require high blood pressure in the cephalopedal area te perform any behaviour, especially feeding, sex, and locomotion, it is hardly a surprise that perforrning these behaviours correlates with increases in heart rate. However, this increase will only be required if the resting heart rate is incapable of generating the required blood pressure to perforrn the behaviour. For this reason the correlation of heart rate wi th BSASs is less than perfecto For example, in Aplysia, food arousal causes no increase in

heart rate unless the animaIs have been previously starved (Dieringer et al., 1978). Why should this be?

Perhaps it is because starvation severely decreases their

resting heart rate (Dieringer et al., 1978)" Once an animal has been 'aroused' by a behaviour such as feeding, its heart rate remains high for sorne time afterwards ( (Dieringer et al., 1978). Therefore it is 'easier' for 177 it to perform other behaviours, as it has the necessary preconditions for any other active behaviour. It may now show a decreased time to respond to other stimuli, but this may be because i t does not need to wai t for bll0d pressure to build up for i t to be capable of responding, and not because a specialized neural system has been activated. For example, Kupfermann and Weiss (1981) found that handling Aplysia increases their responsiveness te food 4 min later. They concluded from this that a 'central arousal' system posi tively medulates a wide variety of behaviours. However, previous to being handled, the animaIs had been quiescent. After being handled the animaIs became 1 aroused ' in the as sense, and had high BSSs. This would have increased their heart rate. The heart rate would still have been elevated even after the animaIs returned to their baseline BSSs 4 min later

(Di.eringer et al., 1978). The faster feeding found by Kupfermann and Weiss (1981) may have been partly due to the residual effects of having an increased BSS on the cardiovascular system. Therefore, although these experiments suggest that a BSCAS exists, they do not necessarily show chat the animal has a special neural system mediating this

'arousal' • Performing any active behaviour has effects on the cardiovascular system and i t may be that i t is 178 l through these effects that behavioural responsiveness is increased. Only time, and more neuroethology, will tell.

Evidence for a BSCAS in molluscs

Because a BSCAS is not the only determinant of behaviour, the correlation between BS and behavioural responsiveness is not likely to be perfecto For example,

Advocat (1980) found that animaIs that had recently been

fed were more active (had a higher BSS) than did starved animaIs, but that the starved animaIs remained more responsive to stimuli, both food and tactile. The observed decrease in responsiveness could be induced

simply by chemostimulation; the ingestion of food was not required. Therefore although BS correlates with

responsiveness (Chapter 5), responsiveness is also

modulated by previous activity. In other words,

mechanisms such as those mediating a decline in feeding

proclivity are also likely to influence the proclivity for other behaviours. This makes conclusive evidence for a BSCAS difficult to glean from behavioural data alone.

The strongest evidence for the existence of such a

BSCAS cornes from experiments by Kupfermann and Weiss

(1982) . By monitoring the output of the metacerebral

cell (MCC), a cell intimately involved in food arousal,

they found that handling the animaIs caused an increase 179 ," in its firing rate. This suggests that a central system that responds to non-food stimuli directly interacts with the feeding specifie arousal system and may be partly

responsible for the increase in food arousal observed in handled animaIs (Kupfermann and Weiss, 1981).

possible mechanisms mediating the BSCAS in Helix

As in Palovcik et al.'s (1982) hypothesis, other workers have suggested that serotonin is invol ved in mediating the BSCAS (leech, Willard, 1981; Aplysia,

Parsons and Pinsker, 1989). However, although 5,7 DHT, a neurotoxin which decreases serotonin levels in Hel isoma

by 85%-90%, did decrease activity in Helisoma for the first 3 days, but the snails showed normal feeding, swimming and mating behaviour long before any significant

recovery in their serotonin levels had occurred (Gadotti et al., 1986). Moreover, very young snails appear to

have no serotonin in their systems, yet these animaIs

still feed (Zakharov and Balaban, 1987) and have typical

daily activity rhythms (personal observations). These

results suggest that serotonin is not a necessary component of the BSCAS in snails. In Helix, BS is associated with physiological

mechanisms that require some degree of coordinated control (Dale, 1973). It is not unrealistic to assume ._---_... _- -_.-

180 that there may be a system in which certain neuraljneuralhormonal units are dedicated to this task. For example, the snail' s emergence from the shell is thought to be controlled, at least in part, by the peptide pQDPFLRF which also increases heart rate (Lehman

and Greenberg, 1987).

Final remarks

As weiss et al. (1981) have pointed out, any motivated behaviour cannot be completely understood until

its underlying physiological basis is elucidated. This will not be possible without clear definitions of the

behaviours involved. Any definition that is so broad as

to encompass phenomena that appear to be distinct even at , ,1 the behavioural level will not be very helpful to .1 1 neuroethologists. Future neuroethological studies

dealing with motivated behaviours would benefit from an

explicit division of the concept of 'arousal' into

proclivity and arousal and from a similar division of the

concept of 'central arousal' into behavioural state

central arousal and behaviour-specific central arousal. The resulting clarity in behavioural terminology will also reduce much of the confusion in the literature

surrounding the concepts and theories concerning (. 'arousal' and 'central arousal' . 181 Bibliography

Advocat, C. (1980) Modulation of defense reflexes in Aplysia californica by appetitive stimulation. Behav. Neural Biol., 28: 253-265. Andrew, R.J. (1974) Arousal and the causation of behavior. Behav., 51: 135-165. Aous, S., Oomura, Y., Yoshimatsu, H. (1988) Neuron activity of the VMH and the MPOA of the female monkey during sexual behaviour. Brain Res., 455: 65-71. Arnold, S.J. and Houck, L.D. (1982) Courtship pheromones: evolution by natural and sexual selection. In: Biochemical Aspects of Evolutionary Biology, ed. M.H. Nitecki, University of Chicago Press, Chicago, pp. 173-211. Audesirk, T.E. (1977) Chemoreception in Aplysia. III. Evidence for pheromone influencing reproductive behaviour. Behav. Biol., 20: 235-243. Audesirk, G.J. and Audesirk, T.E. (1980) Complex mechanoreceptors in Tritonia diomedia: Neural correlates of a change in behavioral responsiveness. J. Camp. Physiol., 141: 111-122. Baerends, G.P. (1976) The functional organization of behaviour. Anim. Behav., 24: 726-738. 182 Bailey, S.E.R. (1981) Circannual and circadian rhythms in the snail Helix aspersa and the photoper iodic control of annual activity and reproduction. J. Comp. Physiol., 142: 89-94.

Balaban, P., & Chase, R. (in press) stimulation of mesocerebrum in Hel ix aspersa inhibi ts the neural

network underlyinq avoidance behavior. J. Comp. Physiol .. Baur, B. (1988) Repeated mating and female fecundity in the simultaneous hermaphroditic land snail Arianta arbustorum. Invert. Reprod. Develop., 14: 197-204. Beach, F.A. (1976) Sexual attractivity, proceptivity, and receptivity in female mammals. Horm. and Behav., 7: 105-138. Bentivegna, F., Cirino, P., Fiore, L. and Gepetti, L. (1988) Control of feeding behaviour in Aplysia depilans by haemolymphatic glucose. Mar. Behav. Physiol., 13: 113-123. Boakes, R. (1984) From Darwin to Behaviourism. Cambridge Univ. Press, Cambridge. Bolles, R.C. (1975) The Theory of Motivation. 2nd ed., Harper and Row, New York, Evanston, London. Bonnefoy-Claudet, R., and Deray, A. (1984) Influence de la durée d' hibernation sur l' activi té reproductrice de l'escargot Helix aspersa. Comptes Rendues de la ( Societe de Biologie de Besancon, 178: 442-449. 1

183 - Bonnefoy-Claudet, R., Deray, A., and Griffond, B. (1987) La reproduction de l'escargot Hel ix aspersa en fonction des conditions d'environnement: donnees synthetiques. Haliotis, 16: 57-67.

Bornchen, M. (1967) Untersuchungen zur Sekretion der

Fingerformigen Drusen von Helix pomatia L .. ~ Zellforschung, 78: 402-426. Brown, R.E. and McFarland, O.J. (1979) Interaction of

hunger and sexual motivation in the male rat: a time sharing approach. Anim. Behav., 27: 887-896.

Cameron, R.A.O. and Carter, H.A. (1979) Intra and interspecific effects of population density on growth

and activity in sorne helicid land snails. ~. Anim. Ecol., 48: 237-246. Chase, R. (1986) Brain cells that command sexual

behaviour in the snail, Helix aspersa. J.

Neurobiol., 17: 669-679.

Chung, O. (1985) An anesthetic for internal operations

on the land snail, Helix aspersa. Veliger, 27: 331-

335. Chung, O.J .0. (1986) stimulation of genital eversion in

the land snail, Hel ix aspersa by extracts of the

glands of the dart apparatus. J. Exp. Zool., 238:

129-139. 184 { Chung, D.J. D. (1987) Courtship and dart shooting behavior of the land snail, Helix aspersa. Veliger, 30: 24-39.

Colebrook, E., Leonard, J. L., and Lukowiak, K. (1984) Physiological evidence for the unitary drive model in

Aplysia. Reg. Pep., 9: 328. Crews, D., & Moore, M.C. (1986) Evolution of mechanisms

eontrolling mating behavior. Sei., 231: 121-125. Croll, R.P. and Chase, R. (1980) Plastieity of olfaetory

orientation to foods in the snail Achatina fulica. Behav. Biol., 19: 261-268. Dale, B. (1973) Blood pressure and its hydraulic

functions in Helix pomatia. J. Exp. Biol., 59: 477- 490.

Dan, N., & Bailey, S.E.R. (1982) Growth, mortality, and feeding rates of the snail Helix aspersa at different

population densities in the laboratory, and the

depression of aetivity of helieid snails by other

individuals, or their mucus. J. Moll. Stud., 48: 257-265.

Davis, W.J. and Gillette, R. (1978) The neural eorrelates

of behavioral plastieity in command neurons in

P1eurobranchaea. Sei., 199: 801-804.

r •

185 Davis, W.J., Mpitsos, G.J., and Pinneo, J.M. (1974a) The behavioral hierarehy of the molluse pleurobranehaea

1. The dominant position of the feeding behavior. ~ Comp. Physiol., 90: 207-224.

Davis, W.J., Mpitsos, G.J., and Pinneo, J.M. (1974b) The behavioral hierarehy of the molluse, Pleurobranehaea II. Hormonal suppression of feeding assoeiated wi th

egg-laying. J. Comp. Physiol., 90: 225-243.

Davis, W.J., Mpitsos, G.J., Pinneo, J.M., and Ram, J.L. (1977) Modification of the behavioural hierarchy of

Pleurobranchaea l. satiation and feeding motivation. J. Comp. Physiol., 117: 99-125.

Dieringer, N., Koester, J., & Weiss, K.R. (1978) Adaptive changes in heart rate in Aplysia californica. J. Comp. Physiol., 123: 11-21. Dorello, P. (1925) Sulla funzione delle glandole digitate nel gen., Helix. Atti R. Accad. Naz. Lineei, 6: 47- 51. Epstein, A.N. (1982) Instinct and motivation as an explanation for complex behavior. In The Physiological Mechanisms of Motivation. ed. D.W. Pfaff, springer-Verlag, New York, Heidelberg, Berlin, pp. 25-58. Everett, R.A., Ostfeld, R.S., and Davis, W.J. (1982) The behavioral hierarchy of the garden snail Helix

aspersa. Z. Tierpsychol., 59: 109-126. \86

Farkas, S.R., and shorey, H.H. (1976) Anemotaxis and odour-trial following by the terrestrial snail Helix

aspersa. ~. Behav., 24: 686-689. Gadotti, O., Bauce, L.G., Lukowiak, K. and Bulloch,

A.G.M. (1986) Transient depletion of serotonin in the

nervous system Helisoma. J. Neurobiol., 17: 431-447.

Gallistel, C.R. (1980) The Organization of Action: A New

synthesis. Lawrence Erlbaum Associates, Hillsdale, New Jersey.

Gillette, M.U. and Gillette R. (1983) Bursting neurons command consummatory feeding behavior and coordinated visceral receptivity in the predatory mollusc

Pleurobranchaea. J. Neurosci., 3: 1791-1806.

Giusti, F., and A. Lepri (1980) Aspetti morfologici ed

etologici dell t accoppiamento in alcune specie della

famiglia Helicidae. IV Congresso della Societa

Malacologica Italiana, Accademia delle Scienze di

Siena detta de'Fisiocritici, pp. 11-71.

Goddard, C.K. (1962) Function of the penial apparatus of

H§lix aspersa. Aust. J. Biol. Sei., 15: 218-232.

Gomot, L., and Deray A. (1987) Les escargots. 19

Recherche, 18: 302-311.

Guemene, D. and Daguzan, J. (1982) Variations des

capacités reprodutrices de l'escargot Helix aspersa

selon son géographique. I. accouplement et ponte.

Ann. Zootech., 31: 391-430. 1

187 - Halliday, T.R. (1983) Motivation. In Causes and Effects. Anl.'mal Behaviour. vol. 1, eds • T .• R Halll.'day a n d P.J.B. Slater, Blackwell Scientific Pub., Oxford, London, Boston, Melbourne, pp. 100-133.

Hamilton, P.V. and Winter, M.A. (1984) Behavioral responses to visual stimuli by the snails Tectarius muricatus, Turbo costanea, and Helix aspersa. Anim.

Behav., 32: 51-57.

Hebb, 0.0. (1955) CNS (Conceptual nervous system). Psychol. Rev., 62: 243-254.

Herzberg, F. (1965) Crowding as a factor in growth and reproduction in Helix aspersa. Amer. Zool., 5: 254. Herzberg, F. and Herzberg A. (1962) Observations on

reproduction in Helix aspersa. Am. MidI. Nat., 68:

297-306.

Hinde, R. A. (1970) Animal Behaviour. 2nd ed., McGraw-

Hill Book Co., New York.

Hogan, J.A. (1980) Homeostasis and behaviour. In:

Analysis of motivational processes, ed. F.M. Toates

and T.R. Halliday, Academie Press, London, New York,

Toronto, pp. 2-21.

Hunt, S. (1979) The structure and composition of the love

dart (gypsobelum) in Helix pomatia. Tissue Cell, 11:

51-61. Hyman, L.H. (1951) The Invertebrates, Platyhelminthes and - Rhynchocoela. vol. 2, McGraw-Hill, New York. , 188 t Ingram, H.M. (1947) The european b~own snail in Oakland, California. Bull. S. Calif. Acad. Sei., 45: 152-159. Jensen, K.R. (1987) Effect of starvation on copulatory activity of Ercolania nigra. Mar. Behav. Physiol., 13: 89-97. Jeppesen, L. (1976) The control of mating behavior in Helix pomatia. Anim. Behav., 24: 275-290. Kandel, E.R. (1976) Cellular Basis of Behavior. W.H. Freeman and Co., San Francisco. Kandel, E.R. (1979) Behavioral Biology of Aplysia. H.H. Freeman and Co., San Francisco. Karlson, P., and Luscher, M. (1959) 'Pheromones': a new term for a class of biologically active substances. Nature, 183: 55-56. Kemenes, G., and S.-Rozsa, K. (1987) The role of serotonergic mechanisms of food induced arousal in the snail Helix pomatia. In: Neurobiology: Molluscan Models, eds. H.H. Boer, H.P.M. Geraerts, and J. Joosse, North Holland Publishing Company, Amsterdam, pp • 277 - 286 • Kendler, H.H. (1987) Historical Foundations of Modern Psychology. Temple Univ. Press, Philadelphia. Kerkut, G.A. and Meech, R.W. (1966) The internaI chloride concentration of H and D cells in the snail brain. Comp. Biochem. Physiol., 19: 819-832. 1... 189 Kerney, M.P. (1976) Atlas of the non-marine mollusca of the British Isles. Institute of Terrestrial Ecology, Cambridge.

Koch, U. and Koester, J. (1982) Time sharing of heart power: Cardiovascular adaptations to food arousal in Aplysia. J. Comp. Physiol., 149: 31-42. Koch, U., Koester, J. and Weiss, K.R. (1984) Neuronal mediation of cardiovascular effects of food arousal in Aplysia, J. Neurophysiol., 51: 126-135. Kohn, J. (1983) Feeding biology in gastropods. In: The Mollusca, vol. 5 Part 2, Physiol., ed. A.S.M. Saleuddin and K. M. Wilbur, Academie Press, Orlando, pp. 1-83. Kovac, M.P. and Davis, W.J. (1980a) Neural mechanism

underlying behavioral choice in Pleurobranchaea. ~ Neurophysiol. 43: 469-487. Kovac, M.P. and Davis, W.J. (1980b) Reciprocal inhibition between feeding and withdrawal behaviors in Pleurobranchaea. J. Comp. Physiol., 139: 77-86. Kravitz, E.A. (1988) Hormonal control of bEthavior: Amines and the biasing of the behavioral output of

lobsters. ~, 241: 1775-1781. Kupfermann, I. (1974) Feeding behavior in Aplysia: A simple system for the study of motivation. Behav. Biol., 10: 1-26. 190 Kupfermann, r., and Weiss, K.R. (1981) Tail pinch and handling facilitate feeding behavior in Aplysia.

Behav. Neural Biol., 32: 126-132. Kupfermann, r., and Weiss, K.R. (1982) Activity of an identified serotonergic neuron in free moving Aplysia

correlates with behavioral arousal. Brain Res., 241:

334-337.

Kuslansky, B., weiss, K.R. and Kupfermann, I. (1987) Mechanisms underlying satiation of feeding behavior

of the mollusc Aplysia, Behav. Neural Biol., 48: 278-

303.

Lee, R.M., and Palovcik R.A. (1976) Behavioral states and feeding in the gastropod Pleurobranchaea. Behav.

Biol., 16: 251-266.

Lehman, H.K. and Greenberg, M.J. (1987) The actions of FMRFamide-like peptides on visceral and somatic muscles of the snail, Helix aspersa. J. Exp. Biol.,

131: 55-68.

Leonard, J.L. and LUkowiak, K. (1984) The behavior of

Aplysia californica I. ethogram. Behav., 98: 320-

360.

Leonard, J.L. and LUkowiak, K. (1986) An ethogram of the sea slug, Navanax inermis. z. Tierpsychol., 65: 327- 345.

( - 191 Lind, H. (1973) The functional significance of the spermatophore and the fate of spermatozoa in the genital tract of Helix PQmatia. J. Zool., 169: 39-64.

Lind, H. (1976) Causal and functional organization of the mating behavior sequence in Helix pomatia.

Behavior, 59: 162-202.

Lloyd, P.E. (1982) Cardioactive neuropeptides in gastropods. Fed. Proc., 41: 2948-2952.

:'ucarz, A. (1984) Etude expérimentale de l'effet du groupement sur la ponte d' Hel ix ?.Spersa. Int. J. Inv. Reprod. Develop., 7: 185-192.

Madec, L. and Daguzan, J. (1987) Etude de quelques facteurs affectant la reproduction de l'escargot

petit-gris, Helix aspersa. Haliotis, 16: 77-91.

McCleery, R.H. (1983) Interactions between activities. In Causes and effects, vol. l, ed. T.R. Halliday and P.J.B. Slater, Blackwell Scientific PUb., Oxford,

London, Boston, Melbourne, pp. 134-167.

McFarland, D., and Sibly, R. (1972) 'Unitary drives'

revisited. Anim. Behav., 20: 548-563.

Meddis, R. (1984) statistics using Ranks - A Unified approach. Oxford: Blackwell.

Meisenheimer, J. (1912) Oie Weinbergschnecke, Leipzig. ( 192 Morand, s. (1988) Elements d'epidémiolgie de Nemhelix baker i (Nematoda, cosmocercidae) parasite de l'appareil genital de Helix aspersa. Haliotis, 18:

297-304. Moulin, O. (1980) Observations sur la reproduction en elevage de l'escargot petit-gris (Helix aspersa) de

France (Normandie) Haliotis, 10: 95-99.

Mpitsos, G.J., and Lukowiak, K. (1985) Learning in gastropod molluscs. In: The Mollusca, vol. 8, Neurobiology and Behavior part 1, ed. A.O.D. Willows,

Academic Press, Orlando, pp. 95-267. Oomura, Y., Aous, S., Koyama, Y., Fujita, F. and

Yoshimatsu, H. (1988) Central control of sexual

behavior. Brain Res. Bull., 20: 863-870.

Painter, S.O., Gustavson, A.R., Kalman, V.K., Naqle, G.T., and Blankenship, J.E. (1989) Induction of copulatory behavior in Aplysia: atrial gland factors mimic the excitatory effects of freshly deposi ted

egg cordons. Behav. Neural Biol., 51: 22-236.

Palovcik, R.A., Basberg B.A., and Ram, J .L. (1982) PJehavioral state changes induced in Pleurobranchaea

and Aplysia by serotonine Behav. Neural Bio1., 35: 383-394.

Parsons, D.W. and Pinsker, H.M. (1989) Swimming in Aplysia brasiliana, behavioral and cellular effects of serotonine J. Neurophysiol., 62: 1163-1176. 193

Peake, J. (1978) Distribution and ecology of stylommatophora. In: The Pulmonates. Vol. 1., ed. V.

Fretter and J. Peake. Academie Press, London. pp. 429-526.

Pentreath, V. W., Berry M. S. and Osborne, N. N. (1982) The serotonergic cerebral cells in gastropods. In: Biology of Serotonergic Transmission, ed. N.N. Osborne, Wiley and Sons, Chichester, New York,

Brisbane, Toronto, Singapore, pp. 457-513.

Potts, D.C. (1975) Persistence and extinction of local populations of the garden snail Helix aspersa in

unfavorable environments. oecologia, 21: 313-334.

Preston, R.J., and Lee, R.M. (1973) Feeding behavior in Aplysia californica: Role of chemical and tactile

stimuli. J. Comp. Physiol. Psychol., 82: 368-381-

Reingold, S.C. and Gelperin, A. (1980) Feeding motor

programme in Limax, II. Modulation by sensory inputs

in intact animaIs and isolated central nervous

systems. J. Exp. Biol., 85: 1-19.

Rivest, B.R. (1984) Copulation by hypodermic injection in

the nudibranchs, Palio zosterae and ,E. dubia

(Gastropoda Opisthobranchia) Biol. Bull., 167: 543-

554. 194 Rosen, s.e., Susswein, A.J., Cropper, E.C., Weiss, K.R., and Kupfermann, I. (1989a) Selective modulation of

spike duration by 5HT and neuropeptides FMRFamide,

SCPb, buccalin, and myomodulin in different classes

of mechanoafferent neurons in the cerebral ganglion

of Aplysia. J. Neurosci., 9: 390-402. Rosen, S.C., Weiss, K.R., GOldstein, R.S. and Kupfermann, r. (1989b) The role of a modulatory neuron in feeding and satiation in Ap1ysia: Effects of

lesioning the serotonergic metacerebra l cells. J.

Neurosci., 9: 1562-1578. Selander, R. K. and Kaufman, D.W. (1975) Genetic structure of populations of the brown snail, Helix

aspersa 1. Microgeographic variation. Evol., 29:

385-401.

Shorey, H.H. (1977) Pheromones. In: How AnimaIs

/: Communicate., ed. T.A. Sebeok, Indiana University 1 '1 Press, Indiana, pp. 137-163.

Sokal, R.R., and Rohlf, F.J. (1981) Biometry., W.H.

Freeman, New York.

Susswein, A.J. (1984) Effects of food depri vation upon

behavioral patterns and time budgeting of Aplysia

fasciata. Behav. Neural Biol., 42: 127-133.

Susswein, A.J., and Bennett, M. V. L. (1979) Plasticity of

feed ing behav ior in the opisthobranch mollusc, Navanax. J. Neurobiol., 10: 521-534. 195 - Susswein, A.J., Gev, S., Feldman, E., and Markovich, S. (1983) Activity patterns and time budgeting of Aplys ia fasciata under field and laboratory

conditions. Behav. Neural Biol., 39: 203-220.

susswein, A.J., Weiss, K.R., and Kupfermann, 1. (1978) The effects of food arousal on the latency of biting

in Aplysia. J. Comp. Physiol., 123: 31-41.

Susswein, A.J., Weiss, K.R. and Kupfermann, 1. (1984) Internal stimuli enhance feeding behavior in the

mollusc, Aplysia. Behav. Neural. BioL, 41: 90-95.

Syzmanski, J.S. (1913) Ein Versuch, die fur das Liebesspiel charakteristischen Korperstellungen und Bewegungen bei der Weinbergschnecke kunstl ich

hervorzurufen. Pfluger's Archiv. Physiol., 149: 471-

482.

Takeda, N. and Tsuruoka, H. (1979) A sex pheromone secreting gland in the terrestrial snail, Euhadra

peliomphala. J. Exp. Zool., 207: 17-26.

Taylor, J.W. (1900) Monograph of Land and Fresh Water Molluscs of the British Isles. I. structural and

General, Taylor, Leeds. Teyke, T., Weiss, K. R. , and KupfeI'lllann, 1. (1990) An identified neuron (CPR) evokes neuronal responses

reflecting food arousal in Aplysia. ~, 247: 85-

87. 196 Toates, F.M. (1980) A systems approach to sexual behaviour. In: Analysis of Motivational Processes, eds. F.M. Toates and T.R. Halliday, Academie Press,

New York, pp. 319-338. Toates, F.M. (1986) Motivational systems. Cambridge Univ. Press, Cambr idge. Toates, F.M. and o 'Rourke, C.O. (1978) Computer simulation of male rat sexual behaviour. Med. Biol.

Enq. comput., 16: 98-104. Tompa, A. S. (1980) The ultrastructure and mineralogy of the dart from Philomycus carolinianus with a brief survey of the occurrence of darts in land snails.

Veliqer, 23: 35-42.

Tompa, A.S. (1982) X-ray radiographie examination of dart formation in, Helix aspersa. Neth. J. Zool., 32: 63-

71.

Tompa, A.S. (1984) Land snails (Stylommatophora). In: The

Mollusca, ed. K.M. Wilbur, vol. 7 Reproduction,

Academie Press, London, pp. 48-140. Trulson, M.E., and Jacobs, B.L. (1979) Raphe unit activity in freely moving cats: correlation with the

level of behavioral arousal. Brain Res., 163: 135- 150.

Tryon, G.W. (1882) structural and Systematic Conchology. vol. 1., The Academy of Natural Sciences of Philadelphia. 197 Tuersley, M.D. (1989) How is food arousal manifested in the pond snail Lymnaea stagnalis?: An overview. !l.:..

Moll. Stud., 55: 209-216.

Tuersley, M.D., and MeCrohan, C.R. (1988) Serotinergic

mod~lation of patterned output in Lymnaea stagnalis.

J. Exp. Biol., 135: 473-486. Tuersley, M.D., and MeCrohan, C.R. (1987) Food arousal in the pond snail, Lvmnaea stagnalis. Behav. Neural

Biol., 48: 222-236.

Van Duivenboden, Y.A. (1984) Sexual behaviour of the hermaphroditie fresh water snail, Lymnaea stagnalis, thesis, vrije Universiteit te Amsterdam.

Van Duivenboden, Y.A., and Ter Maat, A. (1985) Masculinity and reeeptivity in the hermaphrodite pond

snail, Lymnaea stagnalis. Anim. Behav., 33: 885-

891.

Weiss, K. R., Koch, U. T., Koester, J., Mandelbaum, D. E.

and Kupfermann, 1. (1981) Neural and molecular mechanisms of food-induced arousal in Aplysia

californica. Adv. Physiol. sei., vol. 23, Neurobiology of Invertebrates, ed. J. Salanki,

pergamon Press, Budapest, pp. 305-344. 198

\ Weiss, 1eR., Koch, U.T., Koester, J., Rosen, S.C., and

Kupfermann, I. (1982) The role of arousal in modulating feeding behavior of Aplysia: Neural and behavioral studies. In: The Neural Basis of Feeding and Reward, eds. B.G. Hoebel and D. Novin, Haer

Institute, Brunswick, pp. 25-57.

Weygoldt, P. (1977) Communication in crustaceans and arachnids. In: How AnimaIs Conununicate. ed. T.A.

Sebeok, Indiana University Press, Indiana, pp. 303-

333.

Willard, A. L. (1981) Effects of serotonin of t~e generation of the motor programme for swimming by the

medicinal leech. J. Neurosci., 1: 936-944.

Woyciechowski, M. and Lomnicki, A. (1977) Mating frequency between resident and added individuals in a population of land snail Helix pomatia. BulL Acad.

PoL Sei. Biol., Cl. II, 25: 159-162.

Zaferes, A., Skolnick, A. and Tobach, E. (1988) Interindividual contact and copulation in Aplysia

dactylomela, Mar. Behav. Physiol., 13: 221-238.

Zakhorov, I.S. and Balaban, P.M. (1987) Neural mechanisms of age-dependent changes in avoidance

behavior of the snail, Helix lucorum. Neurosci., 23:

721-729.

( 199 - ziv, 1., Benni, M., Markovieh, S., and Susswein, A.J. (1989a) Motivational control of sexual behavior in Aplysia fasciata: Sequencinq and modulation by sexual deprivation and by addition of partners. Behav.

Neural Biol., 52: 180-193.

Ziv, l., Markovien, S. and Susswein, A.J. (1989b) Rules

for behavioral choiee in Aplysia fasicata. ~

Neurosci. Abs., vol. 15 , 509.17.

Zucker, I. (1983 ) Motivation, biologieal elocks, temporal orqanization of behavior. In: Handbook of Behavioral

Neurobiology. vol. 6. Motivation, eds. E. Satinoft

and P. Teitelbaum, Plenum Press, New York, London, pp. 3-21.

An E&gt;L.Amination of the Concept of Arousal Within the Context of The

An E>L.Amination of the Concept of Arousal Within the Context of The